The IT/OT Insider Podcast with David and Willem

In this episode of our podcast, David sits down (again 😀) with Aron Semle, CTO at HighByte, for a follow-up conversation roughly a year after our first podcast together.A lot has changed in a year. When we last spoke, Industrial DataOps was still a concept many people needed convincing about. AI was already making waves, but the link between data foundations and AI readiness wasn’t as obvious to most manufacturing organisations as it is today. So where are we now? And where is the hype outpacing reality?You can listen to the episode, or get the main ideas in this article! Let’s get into it.We Don’t Sell the Problem AnymoreThe market conversation has shifted in the last year. “We don’t sell the problem anymore,” Aron says. “People understand the problem. We have customers and prospects coming to us saying: we know we need to clean up this data to get AI-ready.”That’s a big shift. And it mirrors what we see at the IT/OT Insider as well. During the peak of the LLM hype, everyone wanted AI but nobody wanted to talk about foundations. Now? People come to us and say: we want AI, and therefore we need to build a foundation.That loop is finally closing. And AI is the catalyst.Sebastian from Frost & Sullivan, who presented at HighByte’s DataOps Days late last year, made the argument that AI and DataOps are coupled — hand in hand — and showed the market growing accordingly. He’s right.Thanks for reading The IT/OT Insider! Subscribe for free to receive our posts and support our work.Starting from Where You AreSo what does that foundation look like in practice?Aron breaks it down using a digital maturity lens:If you’ve got equipment with nothing connected and everything is manual, you’re not ready for AI on the factory floor. You need connectivity first — HMIs, SCADA, historians. If you’re at that level, you’re probably ready for the next step: Industrial DataOps.The way HighByte positions it is straightforward. Industrial DataOps is not going to overlap with your existing systems that run the plant. It’s a layer you put on top — one that connects to all those systems, contextualises the data, and provides access to it from other consumers. AI being one of them.The real work, though, is in harmonisation. “How do you consistently represent equipment that could be the same or completely different across facilities?” Aron asks. That’s the scalability piece. At one factory, the production order comes from an ERP or MES. At another, it’s a CSV file on someone’s desktop. If you’re consuming that data from AI — or from anything else — you shouldn’t need to know the difference.UNS: Valuable, But It Was Always Just the StartWe couldn’t have this conversation without touching on Unified Namespace. And Aron’s perspective has matured in the same way the market has.Five years ago, UNS was the start of DataOps at the edge: connect to everything, contextualise it, publish it to an MQTT broker. And for someone coming from nothing — like that bakery with independent machines and local HMIs everywhere — being able to subscribe to a broker and see the real-time state of the factory was genuinely transformative.But now? Aron is candid: “Publishing to MQTT is just the transport of real-time data out. That strategy sticks around.”With AI entering the picture, you also need historical data access, governance across multiple sites and contextual dimensions.“It’s not that anything we did is nulled,” Aron clarifies. “It was the first step of the journey towards something bigger.”LLMs and Agents: Let’s Demystify ThisDavid pushes for definitions, and Aron delivers them with refreshing clarity.* LLMs are the large neural networks most of us experience through chatbots. You put in a request, a response comes out. The context of the chat goes in each time (or a compressed version), and a new response is generated.* Tool calling is when the LLM detects it can’t answer on its own and reaches out to external tools — this is where MCP comes into play.* Agents are really just the code around the LLM. An agent is an LLM in a loop: instead of a human driving the conversation, the agent code handles tool calls, feeds results back into the context window, and decides when the task is done. How autonomous it is depends entirely on the code you write around that loop.“But make no mistake about it,” Aron says. “We’re just building text documents and feeding them into a large language model. And then stuff comes out that’s sometimes in JSON format that we can code against. It’s not beyond that.”LLMs are probabilistic. They’re hallucinating all the time and they just happen to be right a percentage of the time. In manufacturing, where determinism is fundamental, that’s a real constraint.Now imagine that kind of confident hallucination in a pharmaceutical clean room. Or on a production line with multi-million-euro equipment.“If we’re talking about fully autonomous agents that are going to get control of our machines — very bad idea,” Aron says flatly.MCP: Getting Beat Up, But Here to StayMCP (Model Context Protocol) has had a steep rise — and Aron acknowledges it’s getting some backlash. That’s normal for anything with a sharp hype curve.His take: “MCP is remote procedure calling for LLMs.“ It provides discovery, remote calling, and a bit of instruction in between. Where it gets criticised is by technologists who look at it and say: this should just be a REST API.The funny thing? “If you go look at the latest version of MCP, it essentially is a REST API,” Aron says. “They got rid of server-sent events and it’s essentially an API now.”The only scenario where MCP loses relevance, in Aron’s view, is if LLMs get good enough to call conventional APIs directly — and those APIs evolve to be more interaction-oriented rather than CRUD-style developer tools. But for HighByte, it doesn’t matter much either way: you’re still building the Industrial DataOps layer, the contextualisation, the custom pipelines. Whether the AI calls that via MCP or REST is an implementation detail.What matters is that the tools are deterministic. “You define the MCP tool, its inputs, its outputs. If you get a call that’s trying to do SQL injection, you detect that and stop it with deterministic logic,” Aron explains. That’s why HighByte’s implementation was never a static, one-size-fits-all tool set. It’s designed so you build your own tools — like APIs — and control exactly what they do.A Quick Teaser: i3XWe briefly touched on i3X, a new standard that Aron is heavily involved in alongside Jonathan Weiss and Matthew Paris. We won’t go into detail here — we’ll be recording a dedicated podcast with John Dyck and Jonathan Wise in the coming weeks — but the short version is this:Aron sees i3X as a standardised API layer for the factory.Every vendor already has APIs, but everyone does it their own way. If the industry can standardise how contextual data is accessed across vendors, it removes a massive amount of inefficiency.“If cloud vendors step up and build clients for i3X, that is going to be the ingestion highway in and out of the factory,” Aron says. Version one of the spec is expected before the end of Q2 2026.AI for DataOps: The Slider from Manual to AutonomousBeyond using DataOps to feed AI (which is the foundation story), there’s the flip side: using AI to do DataOps faster.HighByte’s latest release includes what they call a Pipeline AI Agent. Pipelines are their ETL tool for moving and transforming industrial data. The agent lets you prompt an LLM to analyse and edit pipeline configurations. What’s clever is the middle ground it occupies. Aron describes it as an AI slider:On one end, fully manual — you know what you’re doing and you want full control. On the other end, full autonomy — let the AI handle it. The Pipeline AI Agent sits in between: human in the loop. You prompt it, the UI shows you inline what it edited, you review, accept or reject, and iterate.This is where AI genuinely accelerates experts. Not by replacing them, but by handling the tedious parts so they can focus on what requires judgement. Give it a P&ID diagram, an Excel sheet, and access to an OPC server, and let it help you get started with contextualisation. Human in the loop to create deterministic output — absolutely needed — but accelerated by AI.Wrapping UpA year ago, we talked about Industrial DataOps as a discipline. Now it’s a recognised market segment. The conversation has shifted from “why do I need a data foundation?” to “how do I build one so AI can actually deliver?”The takeaway from this conversation? The foundation hasn’t changed. The urgency has. And the hype-to-reality gap on agents and autonomous AI is still very wide. The vendors and practitioners who will win are the ones who are honest about where that gap lies — and focus on the use cases that deliver real value today.HighByte will be at Hannover Messe 2026 (April 20-24) — look for Aron and his colleagues at the AWS, Siemens, and Microsoft booths.Thanks for listening — and thanks to HighByte for sponsoring this one!… and if you haven’t listened to the previous conversation we had with Aron, why not do it now?About HighByteHighByte is an industrial software company founded in 2018 in Portland, Maine USA. The company builds solutions that address the data architecture and integration challenges faced by manufacturers and industrial companies as they digitally transform. HighByte Intelligence Hub, the company’s proven Industrial DataOps software, provides modeled, ready-to-use data to the Cloud using a codeless interface to speed integration time and accelerate analytics. The Intelligence Hub has been deployed in more than a dozen countries by the world’s most innovative companies spanning a wide range of vertical markets, including automotive, energy, food and beverage, life sciences, and mining and metals.Stay Tuned for More!🚀 Join the ITOT.Academy (new cohort starts in May!) →Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts:Spotify Podcasts:(*) At the IT/OT Insider we do value our independence and transparency. So as we look for ways to pay the bills we were looking for ways to work with sponsors without giving up on those principles. This is where the idea of sponsors comes from. Together with a few selected sponsors we’ll explore some topics that we both find interesting in the same way we write our normal articles. In the coming weeks you’ll find a couple of pieces that have been sponsored. Feel free to contact us if you are interested in a partnership as well. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

📣 A quick reminder before we start: Our next ITOT.Academy kicks off in May, and our early bird offer is available once more. Want to join our fourth group and learn how to bridge IT and OT? There is no better time than now! 👉 Check the curriculum & enrol via ITOT.Academy 👉If you’ve been following our blog and podcast, you know we spend most of our time in what we call the IT/OT zone: data platforms, connectivity, governance, AI use cases, and everything in between. We’ve also covered the Purdue model, MES, UNS, and even Model Predictive Control. However, we rarely talk about what happens at Level 1 and Level 2 — the actual process control layer that keeps plants running. Not the data it produces. Not the dashboards built on top of it. The control itself.So when we had the chance to sit down with Margret Bauer, Professor of Process Automation at the Hamburg University of Applied Sciences, we jumped at it. Margret is an electrical engineer by training, did her PhD in data analytics on process data back in the early 2000s (before “data analytics” was cool), worked for ABB Corporate Research, and even did early IT/OT integration work — connecting SAP with ABB’s 800xA system back in 2007. (Yes, 2007)PID: The Most Important Algorithm Most Don’t Know AboutLet’s talk about PID control.Not P&ID (the diagram) — PID, short for proportional, integral, derivative. If you studied engineering, you probably had one half-lecture on it, sandwiched between Kalman filters and Lyapunov functions. Easy to overlook.Except it runs the world.Margret was blunt about this: 99.9% of all rockets that have flown into space run on PID control. All the robots you see online? PID underneath. Every valve opening and closing in a chemical plant, a refinery, a bakery? PID.The concept is elegant: the proportional part looks at the present, the integral part looks at the past, and the derivative part looks at the future. Three aspects of time, one controller. As Margret put it: it has the worst name and the best track record of any control strategy out there.But don’t let the simplicity fool you. In practice, PID is hard to implement well. Valves have physical limits — they can’t open beyond 100% (no matter how politely you ask). They take time to respond. And when you need to coordinate two valves for the same flow — say, one big valve for coarse control and a small valve for fine-tuning — the strategies on top of PID get complex fast. These layered strategies exist across every process plant, and they are the strategies that nobody outside the automation world ever talks about.A Dying BreedMargret posted on LinkedIn that process control engineers are a dying breed. When we asked why, her answer was painfully logical: the automation worked. Companies invested in control systems in the 1970s, 80s, and 90s. Plants got more stable. And then management looked at the 20-person controls department and said: “Why do we still need these people? The process runs fine.”So they cut the teams. One by one, across the industry.And that is a major problem.In industry, many control departments are gone — and with them, the expertise to improve or even maintain automation performance. And in academia, process control is barely taught anymore. There are barely any new process control engineers coming through the pipeline. The academics who still focus on it? A handful worldwide, passionate but outnumbered (and Margret surely is passionate 🙂)The AI Reality CheckWillem couldn’t resist: “Margret, of course, I’m going to come in with the solution for all your problems. You need to use AI. It’s going to solve everything.”(We all laughed.)Margret’s response was obviously more measured.One of her master’s students developed a reinforcement learning algorithm for a batch penicillin process that improved throughput by 25%. Genuinely impressive. But it worked because the student had a well-understood simulation model. In the real world? The algorithm wasn’t scalable, wasn’t repeatable, and wouldn’t transfer to another process.This ties straight into something we’ve been discussing a lot recently on this blog: the physical twin problem. AI models need to understand the underlying physics, the process behaviour, the control strategies. Without that, you’re optimising in a vacuum. David’s own experience with nonlinear MPC during his master’s thesis confirmed the same thing — beautiful results on simulated data, useless on real plant data.The takeaway isn’t that AI can’t help. It’s that AI without process knowledge is just maths looking for a purpose.The Operator ParadoxThere’s another angle Margret brought up that resonated with us: the better your automation, the more bored your operators become. One of her students — a former operator — said she used to bring a book to her shift. Press the button, sit down, read for eight hours, hand over to the next shift.That’s great from a stability standpoint. But it creates a dangerous gap. When something does go wrong — and it always does eventually — operators haven’t seen enough upsets to know how to respond.The more you automate, the less exposed your operators are to disturbances, and the harder it becomes to train them for the exceptions.And you can’t just “turn off the MPC layer to make things interesting again,” as David pointed out. So the industry adds another layer — operator training simulators, essentially flight simulators for plant operations. Layer upon layer upon layer.Margret’s view? We’ll never fully automate everything. Every process is different, every plant is an individual. We’ll always need people. The question is how we keep them engaged, trained, and ready for the moments that matter.Why This Matters for the IT/OT WorldIf you’re reading this blog, chances are you’re working on data platforms, digital twins, AI use cases, or integration architectures. All of that is important. But it all sits on top of a foundation that most of us take for granted.Process automation isn’t a solved problem. It’s an under-invested, under-documented, under-appreciated layer that directly determines the quality of the data we work with, the stability of the processes we try to optimise, and the feasibility of the AI models we try to deploy.If the foundation crumbles, nothing above it works.So next time you’re debugging a data quality issue, or wondering why your AI model produces nonsense, or trying to understand why a sensor reading oscillates when it shouldn’t — maybe the answer isn’t in your data platform. Maybe it’s one layer below.Find Margret on LinkedIn: https://www.linkedin.com/in/margret-bauer-a885618/Book ‘Process Control in Practice’ mentioned during the podcast: https://www.amazon.de/Process-Control-Practice-Gruyter-Textbook/dp/3111103722📣 Our next ITOT.Academy kicks off in May, and our early bird offer is available once more. Want to join our fourth group and learn how to bridge IT and OT? There is no better time than now! 👉 Check the curriculum & enrol via ITOT.Academy 👉Stay Tuned for More!Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts:Spotify Podcasts:Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Welcome to the first IT/OT Insider Podcast of 2026! We’re kicking off the year with someone who’s done it all: refineries, equipment reliability, process safety, even the postal industry (and found data at the heart of every role).Kris Doering recently joined SaskEnergy, a government-owned natural gas transportation company in Saskatchewan, where he works on system modelling and asset planning. But before that, he spent years at the Co-op Refinery Complex as superintendent of refinery performance improvement, working on benchmarking, goal-setting, and deploying process safety software. His career also includes stints in equipment reliability, Lean Six Sigma at Canada Post, and early days implementing PI System for upstream gas producers.What ties it all together? Data. And not just collecting it.From Postal Sorting to Refinery BenchmarkingKris’s career path is anything but linear, and that’s precisely what makes his perspective valuable. As he put it:“Data has really been a common thread through the whole career. No matter where I worked, what field I worked in, it’s really been the thing that’s tied all of my roles together.”His time at Canada Post might surprise those who don’t think of postal services as manufacturing. But as Kris explained, the parallels are striking:“You’re getting things off of semi-trailers, you’re sorting mail based on barcodes, you’re dealing with advertising mail, newspapers, parcels from Amazon. There’s a lot of infrastructure and a lot of processes.”Those early Lean Six Sigma projects at Canada Post became foundational for everything that followed. “That work really kind of prepared me for all of the other stuff that I’ve done,” Kris noted.Leading vs Lagging: Why Process Safety Metrics MatterOur conversation centred on process safety. This is a topic that doesn’t always get enough attention outside refineries and chemical plants, but has lessons for anyone working with data and performance management.Kris worked extensively with process safety at the refinery, deploying HSE software and investigating incidents. He explained the critical distinction between leading and lagging indicators: “A lagging indicator is when something bad happens. A leading indicator is something that you can measure that you think will correlate to the outcome.”But here’s where it gets tricky. As Kris pointed out, truly leading indicators—ones that predict future incidents—are extraordinarily difficult to design:“The problem with trying to create a leading indicator for process safety is that, you know, there’s an infinite number of things that could go wrong and an infinite number of conditions that could exist out there.”Instead, what most organisations end up with are proxies—measures of how well they’re managing known risks. And that’s not necessarily a bad thing, as long as you’re honest about what you’re measuring.Front-Line Scoreboards: Making Data Visible Where It MattersAnother practical insight from our conversation was Kris’s experience with front-line scoreboards—physical boards where teams track their own performance metrics.“If you’re tracking the right information and putting it on a scoreboard that is understandable to the people who are doing the work, then those people actually engage with it. They want to know how they’re doing.”This isn’t about surveillance or micromanagement. It’s about giving people the context they need to understand their impact:“They know that they’re there to do a job and they want to know if they’re doing a good job or a bad job... and how to be better at their job.”The key is connecting individual behaviour to outcomes in a way that’s visible and actionable. It’s deceptively simple, but as Kris noted,“Connecting individual behaviour to organisational performance is an inherently complex problem, and replicating it through an organisation is complicated, too.”Complex vs Complicated WorkTowards the end of our conversation, we touched on an important distinction that anyone in industrial operations should understand: the difference between complicated and complex work.Complicated work has known solutions—it might be difficult to execute, but the path is clear. Complex work, on the other hand, involves uncertainty, ambiguity, and problems that aren’t well-defined. As Kris put it:“It’s so important not to complexify things. You must come to the simplest solution. And as you gain more knowledge, more skill, more experience, what ends up happening is you recognise how to make things simple and break things down.”The secret? “A desire to not choose to take on too much for myself.” Sometimes the most skilled move is knowing what not to do 🙂Further ReadingIf you want to dive deeper into some of the topics Kris discussed, here are two excellent resources he recommended:* HSG 254: “Developing process safety indicators - A step-by-step guide for chemical and major hazard industries” Available free at: https://www.hse.gov.uk/pubns/priced/hsg254.pdf* API RP 754: “Process Safety Performance Indicators for the Refining and Petrochemical Industries” Available (subscription required) at: https://www.apiwebstore.org/standards/754 Annex I is particularly recommended for defining process safety data requirements.* The “useless machine”: https://www.cbc.ca/news/canada/saskatchewan/useless-machine-maker-from-regina-gaining-worldwide-fame-1.1326579* And you can find the book “Sooner Safer Happier” by Jon Smart in our Mini Book Library.Stay Tuned for More!🚀 Join the ITOT.Academy (May and September Early birds now available) →Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Today’s guest has lived what most companies are still figuring out: how to turn fragmented systems, manual Excel work, and well-intended “shadow IT” into a coherent Industrial DataOps strategy that actually delivers value.In this episode of the IT/OT Insider Podcast, we sat down with Rija Rakotoarisoa, Group IT Operations & Industry 4.0 Lead at Poclain Hydraulics, a French (international) independent group specializing in the design, manufacture and sale of hydrostatic / electrohydraulic transmissions: motors, pumps, valves, system for off-road or mobile machines and one of the global leaders in hydrostatic transmissions.If you’ve ever found yourself trying to bridge IT and OT while juggling standardization, culture change, and budget cuts… you’ll feel very at home in Rija’s story.From Developer to Industry 4.0 LeaderRija started his career firmly on the IT side: a master’s in computer science, developer turned IT manager, working in a plant where his job was to keep systems running and people connected. Then came the shift.“After five or six years, I felt like I had seen everything. I wanted to do something more than pure IT, something that had a direct impact on the business.”So he went back to school, this time for a master’s in finance. Not because he loved accounting, but because it was his way to “remove the geek tag.”“If you wanted to have more impact, you had to speak the business language.”That change paid off. Rija became both IT and finance manager at one of the company’s plants and learned firsthand what happens when you put technology in service of the business. He used automation to help teams understand their own costs, improve efficiency, and cut the manual data entry that was eating up hours every day.Lessons from Good and Bad ProjectsIn his later roles, including a global Industry 4.0 function, Rija saw dozens of digital projects across multiple plants. Some brilliant, others not so much.“A bad example is when a company rolls out something top-down. They say, ‘This is the strategy, you must implement it,’ without asking the real problems at the plant. It takes time, money, and in the end, nobody uses it.”Sound familiar?The good examples, he says, start from the other direction. From real operational pain points.“When you address the real problem in manufacturing - something that changes the day-to-day of the operational team - then they support you, they use it, and they apply it every day.”It sounds simple, but as he adds, “it’s not.” It takes change management, communication, and people inside each plant who carry the message and help build local momentum.Starting from a Digital GreenfieldWhen Rija joined Poclain Hydraulics, about 6 years ago, it was, as he puts it, “a digital greenfield.” The company had strong IT foundations (infrastructure, networks, ERP), but no consistent support for manufacturing systems yet.“There were many IT/OT projects managed only by operational people. They cared about the end result, but not the implications in term of IT constraints. In the end, you have a big nightmare.”In other words: well-intentioned local initiatives, zero standardization. The kind of environment where every plant has its own version of the truth.So where do you start when the elephant is that big?“We started with the most painful issue: the end-of-line quality control system. Each plant had its own version. We moved from local executable applications to a web-based, centralized one.”Then came work instructions, and so on and so on. It was a classic “bite-by-bite” transformation.How COVID Changed the GameLike many others, Poclain had big plans for a global MES rollout. And then COVID hit. Budgets froze, priorities shifted, and suddenly the grand plan was off the table.“We had to rethink everything. How can we do more with less? How can we use what we already have?”What followed was a shift from “big system thinking” to a more agile, best-of-breed approach.“I always say it’s not a happy event for everyone, but I thank COVID-19,” he laughs. “It forced us to be creative.”That creativity led to the Data Hub project: a pragmatic approach to connecting existing systems, automating data collection, and building live dashboards that operators could actually use.Building a DataOps MindsetThe guiding principle was simple: make data useful, make it live, and make it easy for non-IT users.“I didn’t want my team to be the bottleneck. The system should be usable by non-IT people.”That requirement drove their vendor evaluation which eventually led to selecting Litmus.io as their main Data Hub platform.“Since 2021, we’ve been implementing Litmus as our main data hub. Step by step, we break the silos and build on it.”But technology was only one part of the story. The harder part was governance and culture.“It took a lot of time to explain to top management that the Data Hub is just an enabler. It’s not magic. You need something meaningful for the people at the plants on top of it.”Standardization Without Killing FlexibilityToday, Poclain’s model combines global consistency with local agility.“We master the data model centrally and duplicate it for each site. Plants can adapt the templates locally by defining their equipments and their mappings, but the core remains the same.”The result?Faster rollouts, cleaner data, and dashboards that update automatically without anyone touching Excel.Rija’s model proves that digital transformation doesn’t have to mean disruption, just the right balance between structure and freedom, one data point at a time.Interested in knowing more about Litmus? A few months ago we published our 5 Step Playbook for a Painless DataOps Rollout:And have you already listened to our Industrial DataOps podcast with John Younes?Stay Tuned for More!🚀 Join the ITOT.Academy →Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts:Spotify Podcasts:Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

In our earlier articles, we laid the groundwork for Industrial AI — breaking down the difference between classic AI, generative AI, and agentic AI. But frameworks alone don’t tell the full story. How do these ideas play out when you’re inside a real industrial company, tasked with building teams, getting budget, and making data actually deliver value?For that perspective, we sat down with Nathalie Rigouts, who until recently headed data and analytics at Borealis and is now Head of Business Applications Data and AI at Umicore. Nathalie brings a refreshing, pragmatic voice — someone who moved from finance into IT, and who knows first-hand the reality of building data capabilities in industry.From Finance to Data & AINathalie didn’t start in IT. Her background is in finance, where every month she wrestled with massive spreadsheets just to get accurate actuals. That pain, she recalls, was the start of her data journey:“Every month again, I was struggling with getting the correct actuals. And then of course, you have to make your forecast.”From implementing a financial planning tool, to establishing BI at Borealis, to eventually leading data and analytics, her path shows how close the link is between business need and IT capability. And she’s clear about the lesson: it’s not about technology for its own sake.“It’s not about implementing Microsoft Copilot. You’re not going to gain any sustainable advantage there. But if you can have a deep understanding of the processes in your company, and where data-driven solutions can help, that’s when you start to create value.”Start Small, Sell the SuccessOne of the recurring themes in Nathalie’s story is pragmatism. At Borealis, the team started in 2016 with literally one data scientist and a laptop. “Python notebooks on a laptop, and we started.”The key, she says, is to find enthusiastic allies and solve problems that matter. And once you do, don’t stay modest: market the success internally.“We often forget to sell our success. I would go everywhere and talk about small things we did. And that’s how you gain support for the next steps.”From that first laptop, the team grew, but only because each step came with visible, tangible wins that created pull from the business.Use Cases That MatterSo what are typical use cases in manufacturing? Nathalie sees three common ones:* Predictive maintenance: “If equipment fails often, anomaly detection and predictive maintenance are obvious first steps. But it’s not an easy nut to crack. Often, you don’t have enough failures to feed a model.”* Quality control with computer vision: mainstream, but effective. With enough annotated pictures, good vs bad quality can be classified quickly. The catch? Data Quality.* Logistics optimization: untangling shipping routes and optimizing delivery to customers with AI-based optimization models.These are concrete, valuable problems — and they also highlight the role of data governance. As she recalls with a smile:“We had beautifully annotated data — but all in Finnish. That’s when you realize governance is not optional.”GenAI: Efficiency or Attractiveness?When it comes to Generative AI, Nathalie is cautious. The business case is not always straightforward:“I tried to make the case for Microsoft Copilot. At €30 per user, that’s not small. Does it reduce workforce? No. At best, people spend more time on value-added activities. But what does that bring to the bottom line? Hard to say.”Yet she also sees why companies can’t ignore it.“Companies have to invest in it because it will determine their attractiveness as an employer. New graduates take these tools for granted. If you don’t offer them, you won’t attract talent.”She distinguishes between two levels: workplace efficiency (nice, but hard to quantify) and domain-specific models trained on your own IP. The latter, she believes, is where the real value lies. For example in pharma, where LLMs trained on internal knowledge can speed up R&D. “That’s when AI becomes a true digital co-worker.”Governance, Change, and LegislationOn governance, Nathalie doesn’t mince words:“It’s always the people, the processes, and the tools. The main component around which all of them center is the value case.”Her advice: don’t let your solutions depend on a single enthusiast, and don’t leave an escape hatch back to the old way of working. Change management is part of the job.And on legislation, she takes a positive view:“It’s an opportunity. It forces us to think about awareness, ethics, governance, documentation, monitoring. All things that make sense. Yes, it’s work, but it helps you get budget and build maturity.”Closing ThoughtsWhat we loved about Nathalie’s perspective is how grounded it is. No buzzwords, no silver bullets: just the reality of building teams, solving problems, and learning along the way. Whether it’s predictive maintenance, quality monitoring, or navigating the GenAI hype.Her closing reminder:“Keep it simple, be pragmatic. We built beautiful solutions with just scripting business rules. The business was happy, and nobody needed a fancy machine learning model.”Stay Tuned for More!🚀 Join the ITOT.Academy →Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

When Zev Arnold joined us on the podcast, he brought with him the kind of energy and clarity you rarely get from someone working at the intersection of industry, data, and transformation. As a principal director at Accenture’s Industry X, Zev has spent years working with oil & gas, utilities, mining, and life sciences companies—not just helping them digitize, but helping them make that digitization mean something operationally. “I help engineers and operators use data to improve the way they work,” he said, and that theme stayed with us throughout the conversation.Context is King (But only if the right person owns it)One of the most powerful insights from Zev is his perspective on data contextualization.He tells the story of a compressor engineer who wanted to track starts instead of doing maintenance on a fixed schedule. “Some compressors had start-stop tags, some had rotational speed. The structure of the data needed to support the engineer’s thinking, not the other way around.” That’s when Zev realized: contextualization only works when it’s driven by the user, not imposed by some else. “Give that hierarchy to the compressor engineer and say, this is yours. Own it.”In Zev’s model, self-service is the enabler. If engineers and operators can build their own analytics without writing Python or waiting for a dev team, that’s when transformation becomes real.Platforms that Empower, Not ObstructZev is quick to point out where industrial transformation often stumbles: platforms that weren’t built to scale use cases easily. “We had a platform that worked great for one use case. But every new use case required us to rebuild everything again.”“You want to catch that $50K event before it becomes an environmental incident. The person who understands the problem best is the engineer. We need to give them the tools to act.”The bigger picture? Zev sees a future where operators train and maintain AI systems—even simple expert systems that alert you when a tank overflows. That’s where AI becomes more than a buzzword and actually enters the DNA of industrial work.People, AI, and the Future of WorkZev introduces a compelling framing: people-to-people, people-to-AI, and AI-to-AI. That’s the triangle of future industrial collaboration. A model he borrowed from Paul Daugherty’s book Human + Machine.In this framing, AI isn’t replacing people. Instead, AI becomes part of their toolbox. “Even simple AI—like monitoring sump tank levels—needs someone to train and maintain it,” he says. That job doesn’t belong in a remote digital transformation office. It belongs on the floor, with the engineer who knows the equipment and the impact.Thanks for reading The IT/OT Insider! Subscribe for free to receive new posts and support our work.We just need to catch the real valueIs all this really worth it? Zev answers emphatically, yes. And he points to a hard number: EFORd, the forced outage rate in U.S. power generation. This is a metric used to assess the reliability of thermal power generating units, specifically measuring the probability of a unit being unavailable due to forced outages or deratings when there is a demand for power. It essentially indicates how often a generator is unable to produce power when it's needed. The EFORd rate in the US averages 7.5% (a theoretical 0 value would mean that there are no unplanned outages). If we could close that gap with better decisions, the industry could unlock over $100 billion in value.“And that’s just one industry,” Zev adds. “The ripple effects could be societal: better data centers, climate impact, lower energy bills, even job growth.”Final ThoughtsFrom the subtle distinctions between manufacturing types to the very real, tangible impact of good data and AI done right, we touched it all in this podcast. Whether it’s process or discrete, the message is clear: stop treating transformation like a side project. Get the right tools into the right hands—and let people do what they do best.As Zev put it, “On my worst days, I wonder, is this data really valuable? But 15 years in, I know it is. We just have to use it right.”Stay Tuned for More!🚀 Join the ITOT.Academy →Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts:Spotify Podcasts:Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

In this episode of the IT/OT Insider Podcast, we welcome someone who doesn’t come from cloud platforms, data infrastructure, or connectivity layers. Instead, he brings something equally vital: operational wisdom.Raf Swinnen has spent his career inside factories. From Procter & Gamble to Kellogg's, and later Danone, Raf worked at the intersection of operations and transformation, guiding teams through continuous improvement and later, digital initiatives.What makes his perspective especially valuable? It’s grounded in Lean thinking. Not as a buzzword, but as a real discipline. One that requires a sharp understanding of processes, a respect for people on the floor, and a strong filter for what actually adds value.From Line Leader to Digital Change AgentRaf didn’t start in digital. He started on the floor: managing lines, people, safety, and performance. That experience shaped how he sees digital transformation today: as something that should support operations, not get in the way of them.At Danone, he led digital initiatives at the Rotselaar site (Belgium). The job wasn’t to implement more dashboards. It was to help teams use data to drive better decisions, without losing sight of the fundamentals.“Tech to the Back” — What Digital Should Learn from LeanOne of the most powerful takeaways from this episode is Raf’s principle of “Tech to the back.”“Digital solutions should not be front and center. People and processes should be. Tech should follow.”This is a strong antidote to the over-designed, solution-first approaches that often flood the industrial space. According to Raf, the biggest risk in digital projects isn’t the technology — it’s losing the problem along the way.Three C’s: Connect, Collaborate, and CoherenceAs part of his work with leadership teams, Raf often introduces what he calls the 3 C’s:* Clarity – Where are we going, and why?* Consistency – Are we reinforcing the same messages and systems?* Coherence – Do our tools, apps, and data work together?These are not slogans, they are essential behaviors for any transformation to stick. They also align closely with how we designed the ITOT.Academy, where cross-role learning and shared frameworks are front and center.One of Raf’s biggest contributions came through how he structured teams. In a newly created role as Digital Program Manager, he pulled in both IT and OT voices and even shifted reporting lines to foster true collaboration.He didn’t look for tech wizards. He looked for people with enthusiasm. People who wanted to make a difference. These became his digital ambassadors, key voices from every shift, every team.“When the night shift speaks up, you listen. They see the edge cases nobody else does.”Case Examples: Real Change Starts SmallRaf shared stories from his time at Danone, Kellogg’s, and P&G, where transformation didn’t come from big declarations — but from small, disciplined steps.At one plant, it was about helping teams make better use of their shift handovers.At another, it meant cleaning up data before launching another round of training.At Danone, the challenge was scaling good ideas without flattening local ownership.“Digital without context is noise. The real challenge is creating relevance at the point of use.”Digital with DisciplineRaf’s story is a reminder that digital transformation doesn’t start with technology, it starts with understanding the process. Listening to the people who run it, and designing with clarity and purpose. Whether it's Lean principles, cultural alignment, or simply asking better questions, his approach keeps the focus where it matters: on solving real problems in practical ways.In a time when industrial tech is advancing fast and buzzwords multiply by the day, it’s refreshing to hear someone say: let’s not forget why we’re doing this in the first place.If you’re working in digital, operations, or somewhere in between, this episode is a pause-and-reflect moment.And maybe also a nudge: to push tech to the back, and put people and purpose out front.Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Discover the program and claim your seat here: https://itot.academy 🎙️ In this special episode of the IT/OT Insider Podcast, David and Willem officially announce the launch of the ITOT.Academy!After years of conversations with IT/OT professionals, consultants, and technology vendors, one thing became clear: there’s a huge need for practical, vendor-neutral education to help people work together across IT and OT boundaries.The ITOT.Academy is designed to fill that gap.What you’ll learn in this episode:Why we created the AcademyWho it's for: OT teams, IT teams, consultants, vendorsThe structure of the program: short, live, interactive sessionsWhy it's not about convergence but collaborationWhen the first groups will startHow to sign up and join the first cohorts🚀 Learn more and sign up at https://itot.academy🎧 Subscribe for more honest conversations on bridging IT and OT.Chapters00:00 Introduction to ITOT Academy01:38 Feedback from Subscribers03:47 Target Audience for Training07:24 Training Format and Structure11:19 Core Concepts of the Training13:32 Interactive Sessions and Wrap-Up14:37 Launch Details and Closing Remarks This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Today, we have the pleasure of speaking with Nikki Gonzales, Director of Business Development at Weintek USA, co-founder of the Automation Ladies podcast, and co-organizer of OT SCADA CON—a conference focused on the gritty, real-world challenges of industrial automation.Unlike many of our guests who often come from cloud-first, data-driven digitalization backgrounds, Nikki brings a refreshing and much-needed OT floor-level perspective. Her world is HMI screens, SCADA systems, manufacturers, machine builders, and the hard truths about where industry transformation actually stands today.What’s an HMI and Why Does It Matter?In Nikki’s words, an HMI is:"The bridge between the operator, the machine, and the greater plant network."It’s often misunderstood as just a touchscreen replacement for buttons—but Nikki highlights that a modern HMI can do much more:* Act as a gateway between isolated machines and plant-level networks.* Enable remote access, alarm management, and contextual data sharing.* Help standardize connectivity in mixed-vendor environments.The HMI is often the first step in connecting legacy equipment to broader digital initiatives.Industry 3.0 vs. Industry 4.0: Ground Reality CheckWhile the industry buzzes with Industry 4.0 (and 5.0 🙃) concepts, Nikki’s view from the field is sobering:"Most small manufacturers are still living in Industry 3.0—or earlier. They have mixed equipment, proprietary protocols, and minimal digitalization."For the small manufacturers Nikki works with, transformation isn't about launching huge digital projects. It’s about taking incremental steps:* Upgrading a handful of sensors.* Introducing remote monitoring.* Standardizing alarm management.* Gradually building operational visibility."Transformation for small companies isn’t about fancy AI. It’s about survival—staying competitive, keeping workers, and staying in business."With labor shortages, supply chain pressures, and rising cybersecurity threats, smaller manufacturers must adapt—but they have to do it in a way that is affordable, modular, and low-risk.UNS, SCADA, and the State of ConnectivityNikki also touched on how concepts like UNS (Unified Namespace) are being discussed:"Everyone talks about UNS and cloud-first strategies. But in reality, most plants still have islands of automation. They have to bridge old PLCs, proprietary protocols, and aging SCADA systems first."While UNS represents a desirable goal—a real-time, unified data model accessible across the enterprise—many manufacturers are years (or even decades) away from making that a reality without significant groundwork first.In this world, HMI upgrades, standardized communication protocols (like MQTT), and targeted SCADA modernization become the critical building blocks.The Human Challenge: Culture and WorkforceBeyond the technology, Nikki highlighted the human side of transformation:* Younger generations aren't attracted to repetitive, low-tech manufacturing jobs.* Manual, isolated processes make hiring and retention even harder.* Manufacturers must rethink how technology supports not just efficiency, but employee satisfaction.The future of manufacturing depends not just on smarter machines—but on designing operations that attract and empower the next generation of workers.Organizing a Conference from Scratch: OT SCADA CONBefore wrapping up, we asked Nikki about organizing OT SCADA CON."You need a little naivety, a lot of persistence, and the right partners. We jumped first, then figured out how to build the plane on the way down."OT SCADA CON is designed by practitioners for practitioners—short technical sessions, no vendor pitches, no buzzword bingo. Just real, practical advice for the engineers, integrators, and plant technicians who make industrial operations work.Final ThoughtsIn a world obsessed with the future, Nikki reminds us:You can't build Industry 4.0 without first fixing Industry 3.0.And fixing it starts with respecting the complexity, valuing the small steps, and supporting the people on the ground who keep manufacturing running.If you want to learn more about Nikki’s work, visit automationladies.io and check out OT SCADA CON, taking place July 23–25, 2025.Stay Tuned for More!Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

In this episode of the IT/OT Insider Podcast, where we’re taking a short detour from our usual deep dives into industrial things to explore something broader-but equally vital: how enterprises evolve.We’re joined by Stephen Fishman and Matt McLarty, authors of the book Unbundling the Enterprise, published by IT Revolution. Stephen is North America Field CTO at Boomi, and Matt is the company’s Global CTO. But more importantly for this conversation-they’re long-time collaborators with a shared passion for modularity, APIs, and systems thinking.We’ll talk about the power of preparation over prediction, about how modular systems and composable strategies can future-proof organizations, and-most unexpectedly-how happy accidents (yes, “OOOPs”) can unlock unexpected success.Thanks for reading The IT/OT Insider! Subscribe for free to receive new posts and support our work.From Creative Writing to Enterprise ArchitectureStephen and Matt first connected over a decade ago, when Stephen was leading app development at Cox Automotive and Matt was heading up the API Academy at CA Technologies. Their collaboration grew from a shared curiosity: why were APIs making some companies wildly successful, and why did that success often seem... unplanned?They didn’t want to write yet another how-to book on APIs. Instead, they wanted to tell the bigger story-about why companies who invested in modularity were able to respond faster, seize opportunities more easily, and unlock new business models.“We wanted to bridge the gap between architects and the business. Help tech teams articulate why they want to build things in a modular way-and help business folks understand the financial value behind those decisions.” – Stephen FishmanOOOPs: The Power of Happy AccidentsOne of the big themes in their book is what the authors call OOOPs-not a typo, but an acronym.“Google Maps is the classic story,” Stephen explains. “People started scraping the APIs and using them in ways Google never planned-until they turned it into a massive business. That was a happy accident. And it happened again and again.”So they gave those happy accidents a structure-Optionality, Opportunism, and Optimization.* Optionality: Modular systems open the door to future opportunities you can’t yet predict.* Opportunism: You need ways to identify where to unbundle or where to apply APIs first.* Optimization: Continuously measuring and refining based on real usage and feedback.This framework makes the case that modularity isn’t just a technical preference-it’s a business strategy.Read more about OOOps in this article.S-Curves, Options, and Becoming the HouseAnother concept that runs through the book is the S-curve of growth-the idea that all successful innovations follow a familiar pattern: slow start, rapid rise, plateau, and eventual decline.Most companies ride that first curve too long, betting too heavily on what worked yesterday. The challenge is recognizing when you’ve peaked-and investing in what comes next.“Most people don’t know where they are on the S-curve,” says Stephen. “They think they’re still climbing, but they’re really on the plateau.”That’s where optionality comes in again: the ability to explore multiple futures at low cost, hedging your bets without breaking the bank. They borrow the idea of “convex tinkering”: placing lots of small, low-cost bets with the potential for high upside.“Casinos don’t gamble,” Stephen says. “They set the rules. They optimize for asymmetric value. That’s what this book is trying to teach organizations-how to become the house.”We also wrote about the importance of having cost effective ways to work with data in this previous post:Unbundling is Not Just for Big TechYou might think this is a book for Google, Amazon, or SaaS unicorns-but the lessons apply to every enterprise. Even in manufacturing.“The automotive world has always understood modularity,” Stephen says. “Platforms existed in car design before they existed in tech. When you separate chassis from body and engine, you gain flexibility and efficiency.”And the same applies in IT and OT.* Building platforms of reusable APIs and services* Designing products and processes with change in mind* Investing in capabilities close to revenue, not just internal shared servicesEven internal IT teams benefit from this mindset. Once a solution is decontextualized and reusable, it can scale across departments and generate asymmetric value internally-without needing to sell to the outside world.All Organization Designs Suck (and That’s Okay)A memorable quote in the book comes from an interview with David Rice (SVP Product and Engineering at Cox Automotive):“All organization designs suck”It’s a reminder that there’s no perfect org chart, no flawless model. Instead, success comes from designing your systems, your teams, and your investments with awareness of their limits-and building flexibility around them.“APIs aren’t a silver bullet. Neither is GenAI. But if you design your systems, teams, and investments around modularity and resilience, you’re better prepared for whatever future emerges.”We highly recommend the book Team Topologies as further read on this topic.Final ThoughtsUnbundling the Enterprise is not a technical manual. It’s a mindset. A playbook for organizations that want to survive disruption, scale intelligently, and embrace change-without betting everything on a single future.The ideas in this book are especially relevant for those working on digital transformation in complex industries. It’s not always about moving fast-it’s about moving smart, building for change, and staying ready.You can find the book on IT Revolution or wherever great tech books are sold. And be sure to check out their companion article on OOOPs on the IT Revolution blog.Until next time and stay modular! 🙂Want More Conversations Like This?Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Welcome to the final episode of our special Industrial DataOps podcast series. And what better way to close out the series than with Dominik Obermaier, CEO and co-founder of HiveMQ—one of the most recognized names when it comes to MQTT and Unified Namespace (UNS).Dominik has been at the heart of the MQTT story from the very beginning—contributing to the specification, building the company from the ground up, and helping some of the world’s largest manufacturers, energy providers, and logistics companies reimagine how they move and use industrial data.Every Company is Becoming an IoT CompanyDominik opened with a striking analogy:"Just like every company became a computer company in the ‘80s and an internet company in the ‘90s, we believe every company is becoming an IoT company."And that belief underpins HiveMQ’s mission—to build the digital backbone for the Internet of Things, connecting physical assets to digital applications across the enterprise.Subscribe for free to receive new posts and support our work.Today, HiveMQ is used by companies like BMW, Mercedes-Benz, and Lilly to enable real-time data exchange from edge to cloud, using open standards that ensure long-term flexibility and interoperability.What is MQTT?For those new to MQTT, Dominik explains what it is: a lightweight, open protocol built for real-time, scalable, and decoupled communication.Originally developed in the late 1990s for oil pipeline monitoring, MQTT was designed to minimize bandwidth, maximize reliability, and function in unstable network conditions.It uses a publish-subscribe pattern, allowing producers and consumers of data to remain decoupled and highly scalable—ideal for IoT and OT environments, where devices range from PLCs to cloud applications."HTTP works for the internet of humans. MQTT is the protocol for the internet of things."The real breakthrough came when MQTT became an open standard. HiveMQ has been a champion of MQTT ever since—helping manufacturers escape vendor lock-in and build interoperable data ecosystems.From Broker to Backbone: Mapping HiveMQ to the Capability ModelHiveMQ is often described as an MQTT broker, but as Dominik made clear, it's far more than that. Let’s map their offerings to our Industrial DataOps Capability Map:Connectivity & Edge Ingest →* HiveMQ Edge: A free, open-source gateway to connect to OPC UA, Modbus, BACnet, and more.* Converts proprietary protocols into MQTT, making data accessible and reusable.Data Transport & Integration →* HiveMQ Broker: The core engine that enables highly reliable, real-time data movement across millions of devices.* Scales from single factories to hundreds of millions of data tags.Contextualization & Governance →* HiveMQ Data Hub and Pulse: Tools for data quality, permissions, history, and contextual metadata.* Pulse enables distributed intelligence and manages the Unified Namespace across global sites.UNS Management & Visualization →* HiveMQ Pulse is a true UNS solution that provides structure, data models, and insights without relying on centralized historians.* Allows tracing of process changes, root cause analysis, and real-time decision support.Building the Foundation for Real-Time Enterprise DataFew topics have gained as much traction recently as UNS (Unified Namespace). But as Dominik points out, UNS is not a product—it’s a pattern. And not all implementations are created equal."Some people claim a data lake is a UNS. Others say it’s OPC UA. It’s not. UNS is about having a shared, real-time data structure that’s accessible across the enterprise."HiveMQ Pulse provides a managed, governed, and contextualized UNS, allowing companies to:* Map their assets and processes into a structured namespace.* Apply insights and rules at the edge—without waiting for data to reach the cloud.* Retain historical context while staying close to real-time operations."A good data model will solve problems before you even need AI. You don’t need fancy tech—you need structured data and the ability to ask the right questions."Fix the Org Before the TechOne of the most important takeaways from this conversation was organizational readiness. Dominik was clear:"You can’t fix an organizational problem with technology."Successful projects often depend on having:* A digital transformation bridge team between IT and OT.* Clear ownership and budget—often driven by a C-level mandate.* A shared vocabulary, so teams can align on definitions, expectations, and outcomes.To help customers succeed, HiveMQ provides onboarding programs, certifications, and educational content to establish this common language.Use CaseOne specific use case we’d like to highlight is that at Lilly, a Pharmaceutical company:Getting Started with HiveMQ & UNSDominik shared practical advice for companies just starting out:* Begin with open-source HiveMQ Edge and Cloud—no license or sales team required.* Start small—connect one PLC, stream one tag, and build from there.* Demonstrate value quickly—show how a single insight (like predicting downtime from a temperature drift) can justify further investment.* Then scale—build a sustainable, standards-based data architecture with the support of experienced partners.Final Thoughts: A Fitting End to the SeriesThis episode was the perfect way to end our Industrial DataOps podcast series—a conversation that connected the dots between open standards, scalable data architecture, organizational design, and future-ready analytics (and don’t worry, we have lots of other podcast ideas for the months to come :)).HiveMQ’s journey—from a small startup to powering the largest industrial IoT deployments in the world—is proof that open, scalable, and reliable infrastructure will be the foundation for the next generation of digital manufacturing.If you want to learn more about MQTT, UNS, or HiveMQ Pulse, check out the excellent content at www.hivemq.com or their article on DataOps. Stay Tuned for More!Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Welcome to Episode 10 of the IT/OT Insider Podcast. Today, we're pleased to feature Anupam Gupta, Co-Founder & President North Americas at Celebal Technologies, to discuss how enterprise systems, AI, and modern data architectures are converging in manufacturing.Celebal Technologies is a key partner of SAP, Microsoft, and Databricks, specializing in bridging traditional enterprise IT systems with modern cloud data and AI innovations. Unlike many of our past guests who come from a manufacturing-first perspective, Celebal Technologies approaches the challenge from the enterprise side—starting with ERP and extending into industrial data, AI, and automation.Anupam's journey began as a developer at SAP, later moving into consulting and enterprise data solutions. Now, with Celebal Technologies, he is helping manufacturers combine ERP data, OT data, and AI-driven insights into scalable Lakehouse architectures that support automation, analytics, and business transformation.Thanks for reading The IT/OT Insider! Subscribe for free to receive new posts and support our work.ERP as the Brain of the EnterpriseOne of the most interesting points in our conversation was the role of ERP (Enterprise Resource Planning) systems in manufacturing."ERP is the brain of the enterprise. You can replace individual body parts, but you can't transplant the brain. The same applies to ERP—it integrates finance, logistics, inventory, HR, and supply chain into a single system of record."While ERP is critical, it doesn't cover everything. The biggest gap? Manufacturing execution and OT data.* ERP handles business transactions → orders, invoices, inventory, financials.* MES and OT systems handle operations → machine status, process execution, real-time sensor data.Traditionally, these two have been separated, but modern manufacturers need both worlds to work together. That's where integrated data platforms come in.Bridging Enterprise IT and Manufacturing OTCelebal Technologies specializes in merging enterprise and industrial data, bringing IT and OT together in a structured, scalable way.Anupam explains: "When we talk about Celebal Tech, we say we sit at the right intersection of traditional enterprise IT and modern cloud innovation. We understand ERP, but we also know how to integrate it with IoT, AI, and automation."Key focus areas include:* Unifying ERP, MES, and OT data into a central Lakehouse architecture.* Applying AI to optimize operations, logistics, and supply chain decisions.* Enabling real-time data processing at the edge while leveraging cloud for scalability.This requires a shift from traditional data warehouses to modern Lakehouse architectures—which brings us to the next big topic.What is a Lakehouse and Why Does It Matter?Most people are familiar with data lakes and data warehouses, but a Lakehouse combines the best of both.Traditional Approaches:* Data warehouses → Structured, governed, and optimized for business analytics, but not flexible for AI or IoT data.* Data lakes → Can store raw data from many sources but often become data swamps—difficult to manage and analyze.Lakehouse Benefits:* Combines structured and unstructured data → Supports ERP transactions, sensor data, IoT streams, and documents in a single system.* High performance analytics → Real-time queries, machine learning, and AI workloads.* Governance and security → Ensures data quality, lineage, and access control."A Lakehouse lets you store IoT and ERP data in the same environment while enabling AI and automation on top of it. That's a game-changer for manufacturing."Celebal Tech is a top partner for Databricks and Microsoft in this space, helping companies migrate from legacy ERP systems to modern AI-powered data platforms.There's More to AI Than GenAIWith all the hype around Generative AI (GenAI), it's important to remember that AI in manufacturing goes far beyond chatbots and text generation."Many companies are getting caught up in the GenAI hype, but the real value in manufacturing AI comes from structured, industrial data models and automation."Celebal Tech is seeing two major AI trends:* AI for predictive maintenance and real-time analytics → Using sensor and operational data to predict failures, optimize production, and automate decisions.* AI-driven automation with agent-based models → AI is moving from just providing recommendations to executing complex tasks in ERP and MES environments.GenAI has a role to play, but:* Many companies are converting structured data into unstructured text just to apply GenAI—which doesn't always make sense.* Enterprises need explainability and trust before AI can take over critical operations."Think of AI in manufacturing like self-driving cars—we're not fully autonomous yet, but we're moving toward AI-assisted automation."The key to success? Good data governance, well-structured industrial data, and AI models that operators can trust.Final Thoughts: Scaling DataOps and AI in ManufacturingFor manufacturers looking to modernize their data strategy, Anupam offers three key takeaways:* Unify ERP and OT data → AI and analytics only work when data is structured and connected across systems.* Invest in a Lakehouse approach → It's the best way to combine structured business data with real-time industrial data.* AI needs governance→ Without trust, transparency, and explainability, AI won't be adopted at scale."You don't have to replace your ERP or MES, but you do need a data strategy that enables AI, automation, and better decision-making."If you want to learn more about Celebal Technologies and how they're bridging AI, ERP, and manufacturing data, visit www.celebaltech.com.Stay Tuned for More!Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com

Welcome to Episode 8 of the IT/OT Insider Podcast. Today, we’re diving into real-time data, edge processing, and AI-driven analytics with Evan Kaplan, CEO of InfluxData.InfluxDB is one of the most well-known time-series databases, used by developers, industrial companies, and cloud platforms to manage high-volume data streams. With 1.3 million open-source users and partners like Siemens, Bosch, and Honeywell, it’s a major player in the Industrial DataOps ecosystem.Evan brings a unique perspective—coming from a background in networking, cybersecurity, and venture capital, he understands both the business and technical challenges of scaling industrial data infrastructure.In this episode, we explore:* How time-series data has become critical in manufacturing.* The shift from on-prem to cloud-first architectures.* The role of open-source in industrial data strategies.* How AI and automation are reshaping data-driven decision-making.Let’s dive in.If you like this episode, you surely don’t want the miss our other stuff. Subscribe now! From Networking to Time-Series DataEvan’s journey into time-series databases started in venture capital, where he met Paul Dix, the founder of InfluxData."At the time, I wasn't a data expert, but I saw an opportunity—everything in the world runs on time-series data. Sensors, machines, networks—they all generate metrics that change over time."At the time, InfluxDB was a small open-source project with about 3,000 users. Today, it’s grown to 1.3 million users, powering everything from IoT devices and industrial automation to financial services and network telemetry.One of the biggest drivers of this growth? Industrial IoT."Over the last decade, we’ve seen a shift. IT teams originally used InfluxDB for monitoring servers and applications. But today, over 60% of our business comes from industrial IoT and sensor data analytics."How InfluxDB Maps to the Industrial Data Platform Capability ModelWe often refer to our Industrial Data Platform Capability Map to understand where different technologies fit into the IT/OT data landscape.So where does InfluxDB fit?* Connectivity & Ingest → One of InfluxDB’s biggest strengths. It can ingest massive amounts of data from sensors, PLCs, MQTT brokers, and industrial protocols using Telegraf, their open source agent.* Edge & Cloud Processing → Data can be stored and analyzed locally at the edge, then replicated to the cloud for long-term storage.* Time-Series Analytics → InfluxDB specializes in storing, querying, and analyzing time-series data, making it ideal for predictive maintenance, OEE tracking, and process optimization.* Integration with Data Lakes & AI → Many manufacturers use InfluxDB as the first stage in their data pipeline before sending data to Snowflake, Databricks, or other lakehouse architectures."Our strength is in real-time streaming and short-term storage. Most customers eventually downsample and push long-term data into a data lake."A Real-World Use Case: ju:niz Energy’s Smart Battery SystemsOne of the most compelling use cases for InfluxDB comes from ju:niz Energy, a company specializing in off-grid energy storage.The Challenge:* ju:niz needed to monitor and optimize distributed battery systems used in renewable energy grids.* Each battery had hundreds of sensors generating real-time data.* Connectivity was unreliable, meaning data couldn’t always be sent to the cloud immediately.The Solution:* Each battery system was equipped with InfluxDB at the edge to store and process local data.* Data was compressed and synchronized with the cloud whenever a connection was available.* AI models used InfluxDB data to predict battery failures and optimize energy usage.The Results:* Improved energy efficiency—By analyzing real-time data, ju:niz optimized battery charging and discharging across their network.* Reduced downtime—Predictive maintenance prevented unexpected failures.* Scalability—The system could be expanded without requiring a centralized cloud-only approach."This hybrid edge-cloud model is becoming more common in industrial IoT. Not all data needs to live in the cloud—sometimes, local processing is faster, cheaper, and more reliable."Cloud vs. On-Prem: The Future of Industrial Data StorageA common debate in industrial digitalization is whether to store data on-premise or in the cloud.Evan sees a hybrid approach as the future:"Pushing all data to the cloud isn’t practical. Factories need real-time decision-making at the edge, but they also need centralized visibility across multiple sites."A few key trends:* Cloud adoption is growing, with 55-60% of InfluxDB deployments now cloud-based.* Hybrid architectures are emerging, where real-time data stays at the edge while historical data moves to the cloud.* Data replication is becoming the norm, ensuring that insights aren’t locked into one location."The most successful companies are balancing edge processing with cloud-scale analytics. It’s not either-or—it’s about using the right tool for the right job."AI and the Next Evolution of Industrial AutomationAI has been a major topic in every recent IT/OT discussion, but how does it apply to manufacturing and time-series data?Evan believes AI will redefine industrial operations—but only if companies structure their data properly."AI needs high-quality, well-governed data to work. If your data is a mess, your AI models will be a mess too."Some key AI trends he sees:* AI-assisted predictive maintenance → Combining sensor data, historical trends, and real-time analytics to predict failures before they happen.* Real-time anomaly detection → AI models can identify subtle changes in machine behavior and flag potential issues.* Autonomous process control → Over time, AI will move from making recommendations to fully automating factory adjustments."Right now, AI is mostly about decision support. But in the next five years, we’ll see fully autonomous manufacturing systems emerging."Final Thoughts: How Should Manufacturers Approach Data Strategy?For companies starting their Industrial DataOps journey, Evan has a few key recommendations:* Start with a strong data model → Don’t just collect data—structure it properly from day one.* Invest in developers → The best data strategies aren’t IT-led or OT-led—they’re developer-led.* Think hybrid → Balance edge and cloud storage to get the best of both worlds.* Prepare for AI → Even if AI isn’t a priority now, organizing your data properly will make AI adoption easier in the future."Industrial data is evolving fast, but the companies that structure and govern their data properly today will have a huge advantage tomorrow."Next Steps & More ResourcesIndustrial DataOps is no longer just a concept—it’s becoming a business necessity. Companies that embrace scalable data management and AI-driven insights will outpace competitors in efficiency and innovation.If you want to learn more about InfluxDB and time-series data strategies, visit www.influxdata.com.Stay Tuned for More!Subscribe to our podcast and blog to stay updated on the latest trends in Industrial Data, AI, and IT/OT convergence.🚀 See you in the next episode!Youtube: https://www.youtube.com/@TheITOTInsider Apple Podcasts: Spotify Podcasts: Disclaimer: The views and opinions expressed in this interview are those of the interviewee and do not necessarily reflect the official policy or position of The IT/OT Insider. This content is provided for informational purposes only and should not be seen as an endorsement by The IT/OT Insider of any products, services, or strategies discussed. We encourage our readers and listeners to consider the information presented and make their own informed decisions. This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit itotinsider.substack.com