AI analytics, on your terms, on your infrastructure | AI Analytics Week

Most data teams already have AI running somewhere. The open question isn't whether to allow it — it's how to tighten up governance without becoming the team that slows everyone down. Abed walks through the controls: self-hosting, bringing your own model, keeping data where it already lives, granting access selectively, setting token limits, and auditing what was asked and answered end to end. For people who want AI on, carefully. CHAPTERS: 00:00 – Welcome & housekeeping 01:33 – Why AI governance matters now 07:10 – Access, permissions & cost controls in Metabase 13:35 – Getting AI to answer correctly (Table metadata, Glossary, System prompts) 23:08 – AI usage auditing 26:16 – Bring your own model, self-hosting 31:32 – Q&A 🔷 Event: https://www.metabase.com/events/ai-analytics-on-your-terms-on-your-infrastructure 🔷 Session 1 — AI Analytics, with receipts: https://www.metabase.com/events/ai-analytics-with-receipts 🔷 Session 3 — AI Analytics, with a build button: https://www.metabase.com/events/ai-analytics-with-a-build-button 🔷 AI readiness quiz: https://www.metabase.com/lp/ai-readiness-quiz 🔷 Try Metabase: https://www.metabase.com Join the community: 🔷 https://github.com/metabase 🔷 https://x.com/metabase 🔷 https://www.linkedin.com/company/metabase/ #metabase #dataanalytics #datavisualization #aianalytics