✨ One of the best parts of my role at #TDWI is getting to be in the middle of so many conversations about data, analytics, and AI — from webinars and research interviews to summits, briefings, and informal chats. There’s always something new. So I’m starting a new series: **What I Heard at TDWI Last Week**. Think of it as part notebook, part highlight reel — quick insights and observations from across the TDWI community and our partners. Here are two examples from last week ⬇️ 🔹 In an expert panel on building a data strategy for AI (SAP, Reltio, ZoomInfo): I asked: “What are the top mistakes organizations make when defining and implementing a data strategy for AI?” Three themes stood out: - AI won’t fix bad data. Throwing everything into a large language model won’t magically sort it out. Garbage in still means garbage out. Same issue as big data. Without solid data quality, AI fails. - Scope matters. Success comes from small, measurable outcomes, not boiling the ocean. A focused project makes it easier to engage business owners and security teams. - AI needs its own strategy. It’s not just an extension of your data strategy or another visualization tool. AI requires unique approaches to data, governance, and risk. 👉 Want to hear more? Replay here: https://lnkd.in/e-uW9p-n 🔹 In a webinar with Fivetran and Microsoft: From TDWI’s Data Points survey, I shared that fewer than 10% of organizations consider themselves completely AI-ready. Panelists agreed — structured data may be well understood, but unstructured data is still hard to operationalize. Add brittle architectures and silos, and it’s no wonder business leaders get frustrated. Much like the big data era, companies face familiar challenges, but this time the pace of AI adoption is far faster. 👉 Closing the gap: https://lnkd.in/ex2pqyXw #AI #generativeAI #datastrategy #agenticAI #bigdata #unstructureddata Thanks to panelists: Kelly Kohlleffel, Josh Caplan, Mike Frasca, Brandon Tucker, and Colin D..
