The Ravit Show

I learn the most from people who can explain hard things simply, and my conversation on The Ravit Show with Patricia Moore, AI Field CTO at Boomi, was one of those at Boomi World 2026.Patricia was direct about why so many AI agent pilots stall. Most teams rush to deployment before doing the work on context, data readiness, and governance. That is the gap between the enterprises getting real value and the ones still running experiments.We spent real time on context engineering. Everyone uses the phrase, but very few can explain it. Patricia made it obvious why context is not just another feature. It is what decides whether an agent can be trusted inside a real business. The same thinking applies to hallucinations. The fix is not bigger models. It is better grounding, cleaner data, and tighter checks around the agent.The part I enjoyed most was her view on the shift happening inside enterprises moving from experiments to real outcomes. The leaders getting it right are not chasing AI for the sake of AI. They are tying every initiative to outcomes that actually matter.My takeaway. The companies winning with agents are treating context, governance, and outcomes as the real product. The model is just one part of the story.#data #ai #BoomiWorld #theravitshow #BoomiAmbassador

One of the sharpest architecture conversations I had at Boomi World 2026 on The Ravit Show was with Kenneth Maglio, Principal Architect at World Wide Technology. His view on agentic AI was refreshingly honest. Can't wait to do this again!!!!A lot of teams are still debating whether to prioritize data management or agentic AI. Ken's answer was simple. That debate is the problem. If you separate them, you end up with agents that look impressive in a demo and fall apart in production.We talked about what his team was spending too much time on before Boomi, and how much of that work was not moving the business anywhere. The bigger shift was around data freshness. Ken made the case that this is the single biggest factor in whether an agent can actually be trusted. Stale data is not a small issue. It is the difference between a system that scales and one that quietly erodes confidence across the business.We also got into the measurable outcomes WWT has seen since adopting the Boomi architecture, and what would break if that layer was removed. His answer made it clear how foundational this has become.The takeaway for me. Agentic AI in the enterprise will not be won by whoever has the best model. It will be won by whoever has the cleanest, freshest, most governed data feeding those agents in real time.#data #ai #BoomiWorld #theravitshow #BoomiAmbassador

Enterprise software is changing. I sat down with Brian Landsman, CEO of AgentExchange at Salesforce, to talk about what an agent-first future actually looks like. #salesforcepartnerThis wasn’t a surface-level conversation. We went deep into what’s coming next.Here’s what stood out:* AgentExchange evolved from a marketplace directory into a commerce and discovery layer, is rethinking how enterprises deploy software* Headless architectures could fundamentally reshape how people interact with enterprise systems since traditional UIs matter less* Agents are moving from assistants to becoming the primary interface* Workflows need to be redesigned from the ground up for an agent-first world* Success will be defined by agents executing end-to-end tasks, not just supporting humans* The gap between AI pilots and production is finally starting to closeWe also discussed how individuals can go to market faster with $50M AgentExchange Builders Initiative.Watch the full conversation and let me know what you think.#data #ai #tdx26 #salesforce #workflows #api #headless360 #agentexchange #apps #theravitshow

300 to 600 hours reclaimed every single week. Ticket creation cut by 75%.These are not projections. This is what DocuSign is actually seeing from Atlassian's Rovo rollout right now.New episode of The Ravit Show is live with Shivi Singh Verma, MBA, PMP®, CSM®, PMI-ACP®, ITIL®, Senior Manager of Engineering at Docusign, recorded at Team '26 in Anaheim.If you have been waiting for an enterprise AI deployment story that goes past pilots and demos, this is the one to watch.Shivi leads GenAI and AI Agentic strategy at DocuSign. They have actually done the hard work most companies are still talking about. Phased rollout, real guardrails, measured ROI, and a clear plan for what comes next. Their philosophy on this is sharp: adopting AI at scale requires foundational trust, robust governance, and clear guardrails. Not optional, not later, on day one.What we got into:- The tipping point. What finally convinced DocuSign to move forward with Rovo. There is a specific moment Shivi described that I think every engineering leader weighing this decision needs to hear.- The phased rollout. What the pilot looked like, what surprised Shivi as they expanded beyond it, and the guardrails they put in place that they would recommend to other enterprises starting today. This is the playbook section.- How they actually measured ROI. Most companies struggle to prove AI value to leadership. DocuSign did not. I asked Shivi how they measured the 300 to 600 hours weekly and the 75% ticket reduction, and what convinced their leadership these gains were real and sustainable. The answer is more disciplined than I expected.- What comes next. DocuSign is planning to let non-technical teams build their own governed agents through Rovo Studio, and shift from reactive AI to proactive AI. We spent time on what that future looks like, and what they are doing now to prepare for it.- The line from Shivi that stayed with me: AI at enterprise scale is not a model problem. It is a trust problem. Get the governance right first and the productivity gains follow. Skip that step and the project will not survive its first incident.If you are an engineering leader, a CIO, or anyone trying to build the business case for enterprise AI inside your own company, watch this one. Shivi gives you the playbook.Big thank you to Shivi for the openness about what worked and what was harder than expected. And to the Atlassian team for the front-row access at Team '26.#data #ai #atlassian #team26 #theravitshow

What happens when the database becomes an active participant in AI applications instead of just a place to store data? In this session of The Ravit Show, I sat down with Jay Gordon and Patty Chow to unpack the biggest announcements and takeaways from CosmosDB Conf!!!!One theme stood out throughout the conference:AI is not just changing applications. It's changing the database itself.We discussed:- How OpenAI scales from zero to millions of queries per second- Why Walmart relies on globally distributed architectures to keep checkout systems running during failures- How vector search, full-text search, and hybrid search are becoming native database capabilities- The rise of agent memory architectures and AI-native applications- Why developers need real-time visibility into query costs- How to think about CosmosDB vs Azure DocumentDB based on workload requirements- What the Azure CosmosDB Agent Kit means for developers building AI-powered systemsOne of my biggest takeaways was that retrieval is increasingly moving into the database layer itself. Instead of stitching together multiple services, developers can now work with a more unified approach to search, AI, and data.If you're building AI applications, working with data infrastructure, or trying to understand where databases are headed next, this conversation is worth watching.The full interview is now live.What was your biggest takeaway from CosmosDB Conf this year?#data #ai #azure #cosmosDB #microsoft #api #microservices #theravitshow