The Ravit Show

At AI4, I spoke with Kevin McGrath, Co-Founder and CEO of Meibel, on The Ravit Show about the hard truths behind the numbers: nearly 80% of AI projects never make it to long-term success.Kevin explained that it often comes back to data. Customers run into incomplete or poor-quality datasets, or they cannot unify data across silos. Without fixing those fundamentals, even the most ambitious AI initiatives will falter.We also explored why AI demos that wow investors often stumble when they reach real customers. Scaling is the breaking point; infrastructure, workflows, and production realities expose weaknesses that prototypes cannot hide.With GenAI, Kevin sees teams hitting roadblocks around context, reliability, and safety. Building proofs of concept is straightforward, but building systems that are trustworthy and scalable is the real challenge.The takeaway: AI success depends less on the demo and more on how well you prepare data and infrastructure for real-world scale.#data # ai #agents #meibel #theravitshow

AI is only as powerful as the data you give it. At AI4 this week, I sat down with Zack Wenthe 🧠📢, Director of Product Marketing at Tealium, to unpack why a trusted, real-time data foundation is becoming the make-or-break factor for AI success.We started with the basics — trust. If your AI does not have reliable, clean data, everything it produces is at risk. Zack shared why this foundation is non-negotiable for enterprises.Then we explored “real-time.” AI has redefined it from milliseconds to near-instant context and action. Tealium’s approach to orchestrating customer data in real time is not just about speed — it is about making the right decision at the right moment.Zack also gave a preview of his talk, 5 Trends That Will Change Your Customer Data Strategy & Marketing Future. His point was clear: traditional strategies cannot keep up, and marketers who embrace AI-driven approaches will be the ones to thrive.We closed on action. From AIStream and MCP integration to partnerships with Databricks and Snowflake, Tealium provides practical ways to start leveraging AI today — with innovations like Sia pushing the boundaries even further.The takeaway: In the AI era, your data foundation is your competitive advantage.#data #ai #ai4 #tealium #theravitshow

The AI ecosystem is moving fast and that includes the companies behind it. I spoke with Jeff Mills, President of iMerit Technology on The Ravit Show about what recent founder exits and shifting strategies from big players mean for the data annotation and labeling industry!!!!Jeff explained that in a market where trust and longevity matter more than ever, companies building high-stakes AI products cannot afford to rely on unstable or short-term partners. The risk is not just disruption, it is potential failure of entire AI initiatives if the data foundation is compromised.We also discussed what AI companies should look for in a data quality partner today. With innovation moving at breakneck speed, reliability, scalability, and domain expertise are just as critical as technology. It is not about one-off projects but about building trusted, long-term partnerships that evolve with the industry.Jeff highlighted what sets iMerit apart from the dozens of vendors in the market: a proven ability to deliver high-quality data pipelines at scale, a focus on trust, and a commitment to supporting customers as their needs shift from simple annotation toward powering complex AI systems.The takeaway: as AI advances, data quality remains the foundation. And choosing the right partner may determine not only the success of your current models, but your ability to innovate sustainably in the years ahead.What do you think?#data #ai #imerit #ai4 #theravitshow

The era of traditional digital transformation is over. Is it? What do you think?I spoke with Tim McIntire from Hyland on The Ravit Show about why enterprises can no longer think of transformation as just migrating to the cloud or modernizing applications. Those moves may reduce costs or simplify IT, but they do not fundamentally change how businesses operate!Tim explained that the real shift comes with AI agents. These agents are not just tools that automate repetitive tasks. They are intelligent systems that can learn, adapt, and work alongside people in real timeThey change the way decisions are made, how information flows, and how quickly businesses can respond to customers and markets.We also talked about why cloud migration alone is not enough. Without intelligence embedded into processes, enterprises simply recreate the same inefficiencies in a different environment. True transformation requires agents that can cut through friction, connect data, and surface insights the moment they are needed.Hyland calls this agentic transformation, a continuous evolution where AI agents are woven into the fabric of business operations. It is not a one-time project, but an ongoing journey where processes become more adaptive, collaborative, and intelligent over time.The key takeaway: in the AI era, transformation is no longer about where your systems run, but how intelligently they operate and how seamlessly they work with the people who rely on them.#data #ai #agents #hyland #theravitshow

I spoke with Tim McIntire, EVP and CTO, Hyland on The Ravit Show about one of the biggest questions enterprises are facing today: how to scale AI agents responsibly and effectively!!!!Tim shared that the real challenge is not just building the agents, but preparing the right foundation for them to operate at scale. Without trusted data and thoughtful governance, AI agents risk creating more problems than they solve.Hyland is working with customers to put that foundation in place, ensuring AI agents are embedded into workflows in a way that augments people rather than replaces them. The focus is on collaboration between humans and agents, not competition.Looking ahead, Tim sees AI agents becoming much more integrated into day-to-day operations within the next one to two years. Success will come to those enterprises that can strike the right balance: equipping agents with data they can trust, while maintaining transparency and control for the humans who rely on them.The future of AI is not just about agents running faster, it is about building systems where people and AI can work better together.I can't wait to meet again at the COMMUNITYLIVE next week in Vegas!!!!#data #ai #ai4 #hyland #theravitshow

Everyone’s excited about GenAI — but building real, reliable agents? That’s where it gets tricky. I chatted Holly Smith and Nick Karpov from Databricks at the Data + AI Summit on The Ravit Show to talk about the tools that are actually making agentic AI work in the enterprise.We covered two big launches:🔹 Agent Bricks – Designed to reduce the cost and complexity of deploying AI agents in production. Holly and Nick walked me through how it’s solving real-world problems — and what kinds of use cases are already gaining traction across industries.🔹 MLflow 3.0 – The latest update has been reimagined for GenAI workflows, with better support for LLMs and more streamlined experimentation. It’s a big step toward faster iteration and safer deployment of AI models.We also geeked out a bit on how Databricks uses its own platform internally — always fun to hear how the builders are also the power users.Great convo with two people who live and breathe this stuff.#data #ai #databricks #DataAISummit #theravitshow

How do you modernize data management while keeping the reliability of the mainframe?This was the central theme of my conversation with Tomer Zelberzviq, Senior Product Manager at BMC Software. Mainframes have been the backbone of enterprise IT for decades, but today’s cloud-driven environment demands scalability, cyber resiliency, operational efficiency, and compliance at a whole new level.In this interview, Tomer and I discussed:-- Why modernization is still a major challenge for enterprises running on mainframes-- How technologies like Cloud Data Sets are changing the game for security, resilience, and cost efficiency-- Strategies to protect mainframe data against ransomware and cyber threats-- How to make mainframe data accessible for AI and analytics, eliminating silos without disrupting core systemsThis session goes beyond buzzwords. It’s about real solutions that help enterprises bridge legacy reliability with modern innovation.#data #ai #mainframes #bmc #theravitshow

AI doesn’t work without the right data layer. And that’s why Lakebase is turning heads this year. I had a blast chatting with Holly Smith and Nick Karpov from Databricks at the Data + AI Summit, where we dug into some of the most exciting (and most technical) announcements from this year — especially Lakebase.We unpacked why Lakebase is such a big deal:• It’s not just another feature — it’s about making the database layer finally work for AI.• Traditional databases weren’t built for the scale, speed, or flexibility that production AI agents demand — Lakebase is changing that.• We also talked about the Neon acquisition — and why it’s generating so much buzz in the data community.From a data engineer’s lens, the big question is always: Does this make my job easier?Lakebase, combined with AI-native platform features, is a big step toward faster, cleaner AI project delivery.Holly and Nick also gave me a look at what’s coming with Spark Declarative Pipelines — bridging declarative simplicity with real developer control.Big thanks to both for the great chat.#data #ai #DataAISummit #databricks #theravitshow

Privacy, governance, and security — three things that are becoming even more critical as AI goes mainstream. I spoke to Anshu Sharma, Co-Founder & CEO of Skyflow, at Data + AI Summit on The Ravit Show to talk about where the market’s heading and how enterprises are thinking about secure AI adoption.We covered:🔹 The latest trends in data governance and privacy, especially in the context of GenAI🔹 What’s getting enterprises excited right now — and where they’re putting their focus🔹 How this space is likely to evolve in the next 3–6 months, as more orgs go from experimentation to production🔹 And Skyflow’s long-term vision: enabling innovation without compromising trustIf your AI strategy includes sensitive data (and it probably should), this conversation is worth a listen.#data #ai #DataAISummit #databricks #skyflow #governance #security #theravitshow

I had a chance to interview one and only Jeff Lunsford, CEO of Tealium, at the Data + AI Summit, and we had a real conversation about what it actually takes to deliver AI-powered customer experiences. One key takeaway: AI is only as good as the data feeding it — and that data needs to be real-time, consented, and privacy-compliant. That’s where Tealium comes in.We spoke about the launch of CloudStream and how it’s helping enterprises push clean, event-driven customer data into platforms like Databricks without delays or duplication.Jeff also shared how Tealium’s working with Databricks to close the gap between marketing data and AI — making it easier for teams to build smarter, faster, and more responsible experiences.Always great to dig into how different layers of the stack come together.#data #ai #tealium #DataAISummit #agents #cdp #databricks #theravitshow

One of the most honest conversations I had at Data + AI Summit by Databricks was with Sally Liu (Machine Learning Engineer at Kumo) and Raymond Yao (Coinbase) — two people building and applying AI where it really matters. Learn more -- https://kumo.ai/?utm_source=ravit&utm_medium=influencer&utm_campaign=snowflake_databricks2025&utm_term=the_ravit_showWe talked about:- The biggest opportunities and roadblocks in deploying machine learning at scale — across both platforms and real-world fintech- What’s getting their attention beyond LLMs — including models and infrastructure they believe are next- The real talk on ROI — not every AI project pays off, so we dug into which efforts have actually delivered business impact- And that eternal question: build vs. buy? Their perspectives on when to lean on in-house innovation vs. external tools were refreshingly groundedIt’s always great to hear from folks who are in the trenches, balancing experimentation with results.#data #ai #DataAISummit #databricks #kumo #agents #machinelearning #theravitshow

Clean data isn’t just nice to have — it’s the difference between insight and noise. At Databricks’ Data + AI Summit, I had a conversation with Sal Pecoraro about how Data Axle is helping enterprises make their data not just accessible, but actionable.We covered a lot of ground, including:- Their new presence on the Databricks Marketplace, and what that means for seamless access to high-quality, enriched datasets- Why data quality is central to any CDP initiative — and how it directly impacts customer understanding and segmentation- The move to first-party data strategies, and how companies can maximize the value of what they already own- The evolution of data marketplaces and why they’re becoming a key pillar in modern data architecture- Misconceptions around data enrichment, especially when it comes to scale and relevanceSal brought clear, grounded advice that any data leader should pay attention to. #data #ai #DataAISummit #databricks #dataaxle #agents #identity #theravitshow

Customer identity is more than a profile — it’s a foundation.I recently chatted with Marc Sabatini at Databricks’ Data + AI Summit to unpack why identity resolution is fast becoming the backbone of effective marketing strategies.Marc introduced me to the concept of the “Identity Spine” — a unified thread that connects all customer interactions and behaviors. In a world of rising privacy expectations and the fall of third-party cookies, this kind of consistency is more valuable than ever.We dug into:- How a strong Identity Spine helps unify customer data and drive real ROI- Why so many transformation projects stall, and what it takes to get them back on track- What it means to actually activate identity across an enterprise, not just store it- Emerging martech trends for the next 12–18 months — and how Data Axle is preparing for what’s nextGreat perspective from someone who’s seen the strategy behind the buzzwords. Watch the full interview here#data #ai #DataAISummit #databricks #dataaxle #agents #identity #theravitshow

Had a great conversation with Eric Pité, SVP of Product & Strategy at Posit PBC (formerly RStudio) on The Ravit Show, at the Data + AI Summit by Databricks.We talked about:-- The evolution from RStudio to PositIt’s more than just a name change — it’s a signal of how seriously they’re embracing Python while staying true to their R roots and open-source mission.-- Open source meets enterpriseEric shared how Posit is navigating the balance between community contribution and commercial sustainability — and why that mission still matters in an AI-first world.-- Positron + DatabricksTheir strategic partnership is one to watch. Positron is making it easier to do high-quality data science (in R and Python) inside the Databricks ecosystem — with an emphasis on collaboration, reproducibility, and performance.-- AI’s place in modern data scienceWe chatted about how Posit sees the role of AI in helping data teams be more productive — without losing rigor or transparency.-- The power of storytelling with codeOne of my favorite parts: the emphasis on communication. It’s not enough to build models — data scientists need to share insights clearly, and Posit is building for that.Big thanks to Eric for the thoughtful chat — and to the Posit team for continuing to push the boundaries of open, collaborative data science.#data #ai #DataAISummit #Databricks #cdata #agentic #connectors #theravitshow

Every year at Data + AI Summit, there’s a shift in energy — and this year, it’s clear: AI is in production, and the spotlight is finally on the data layer. Caught up with my friend Raj Bains, CEO & Co-Founder of Prophecy on The Ravit Show, to talk about what’s changed — and how their mission to combine self-service with governance is resonating more than ever.We covered:-- Why “democratizing AI” isn’t just a nice phrase — it requires a platform that prepares the room with the right guardrails-- Real-world traction with JPMC and Amgen — and how success turned into investment-- What it really takes to build a production-ready, governed data pipeline in under 10 minutes (hint: it's not just speed, it's control)-- Their early agentic AI features helping analysts find and transform data with smarter guidanceRaj also shared what’s changed in the past year — a clear push across customers to modernize, consolidate legacy tools, and go deeper with AI features.Lots of energy around Prophecy this year — and for good reason.#data #ai #DataAISummit #Databricks #prophecy #agentic #connectors #theravitshow

How do you modernize data management while keeping the reliability of the mainframe?This was the central theme of my conversation with Tomer Zelberzviq, Senior Product Manager at BMC Software. Mainframes have been the backbone of enterprise IT for decades, but today’s cloud-driven environment demands scalability, cyber resiliency, operational efficiency, and compliance at a whole new level.In this interview, Tomer and I discussed:-- Why modernization is still a major challenge for enterprises running on mainframes-- How technologies like Cloud Data Sets are changing the game for security, resilience, and cost efficiency-- Strategies to protect mainframe data against ransomware and cyber threats-- How to make mainframe data accessible for AI and analytics, eliminating silos without disrupting core systemsThis session goes beyond buzzwords. It’s about real solutions that help enterprises bridge legacy reliability with modern innovation.#data #ai #mainframes #bmc #theravitshow

I caught up with Rahul Pahuja from CData Software to dive into how they’re helping enterprises unlock more value from their data inside Databricks — especially as companies push harder into AI and agentic workloads.Here’s what stood out from our conversation:🔹 Enhancing the Lakehouse for AIWhile Databricks positions itself as the Lakehouse for AI, Rahul shared how CData extends its reach by connecting Databricks to hundreds of live data sources — no heavy pipelines, no data duplication. That’s a game-changer for teams needing immediate access to siloed enterprise data.🔹 Activating the 70% of underutilized dataWith a large chunk of enterprise data still going unused for analytics, CData is stepping in to make that data accessible and actionable — all within the Databricks environment. It’s not just about integration, but about activation.🔹 Accelerating agentic workloadsWe also discussed the rise of agent-based systems. Rahul explained how CData helps feed real-time, dynamic data into these AI agents, allowing them to act and respond faster — and with better context — inside Databricks.🔹 Real-world tractionFrom major enterprises slashing integration timelines to faster AI-driven outcomes, Rahul shared success stories that really bring their value to life. The numbers were compelling.If Databricks is your AI foundation, CData could be the unlock to power it with all your enterprise data — fast.Appreciate the insights, Rahul — and hats off to the CData team!#data #ai #DataAISummit #Databricks #cdata #agentic #connectors #theravitshow

Spoke with Amit Sharma, CEO of CData Software, at the Data + AI Summit on The Ravit Show, and we dove into a critical but often overlooked topic: data connectivity for AI agents. With the rise of agentic AI, the old ways of integrating data just don’t cut it. Amit broke down why real-time, context-aware access is now a must — and how CData is helping lead that shift.We talked about:-- The Model Context Protocol (MCP) — a powerful new way to serve up enterprise data to LLMs and agents without building brittle pipelines-- Why CData’s approach to live connectivity is a better fit for GenAI use cases than traditional ETL-- Where agents are being adopted first, based on early traction with their MCP servers-- And the biggest misconception he sees: thinking that LLMs can work well without solving the data access problem firstIf you’re thinking about scaling AI agents across your enterprise, this conversation is worth a listen.#data #ai #DataAISummit #Databricks #cdata #agentic #connectors #theravitshow

What does it really mean to be “agentic-ready”? On The Ravit Show, I had an insightful conversation with Aakarsh Kishore, Chief Product Officer at MathCo, at the Data + AI Summit, and we unpacked what that looks like in practice.We explored:-- Contextual solutions — and why understanding enterprise context is key to building AI agents that actually deliver value-- The concept of a “supercharged brain” — MathCo’s vision for decision intelligence that goes beyond dashboards and reports-- NucliOS — MathCo’s AI decisioning platform that plugs into the Databricks Data Intelligence Platform to help enterprises scale decision-making across teamsAakarsh also shared how the MathCo + Databricks partnership is shaping real-world use cases — and how they’re thinking about AI that’s both scalable and deeply enterprise-aware.#data #ai #databricks #DataAISummit #theravitshow

I had a great conversation with Steven Touw, CTO of Immuta, at the Databricks Data + AI Summit — diving deep into one of the most overlooked challenges in scaling AI: data access.We started with a key question — why are traditional request-and-approve workflows no longer working for modern data teams?Steve broke it down: with more users, more data sources, and faster decision cycles, manual processes create friction, slow teams down, and often lead to over-permissioning or shadow access.We explored how policy-based access control is emerging as the scalable, secure alternative — automating who gets access to what, under which conditions, in real time.Steve also shared how Immuta is helping enterprises balance flexibility and control — enabling faster AI development without compromising on governance, compliance, or data privacy.This shift in thinking is critical for any company looking to unlock the full value of their data while keeping it safe.Appreciate the insights, Steve!!!!#data #ai #DataAISummit #Databricks #immuta #agentic #connectors #theravitshow