DataHub July 2026 Town Hall: Real Stories in Production feat. Miro & Just Eat Takeaway
Scale, Context, and Community. This town hall is all about DataHub in production: different teams, different problems, one platform underneath them all. Hear how Miro lifted their AI agent's text-to-SQL accuracy from under 40% to over 90% using metadata signals in DataHub, how Just Eat Takeaway runs DataHub self-hosted at scale, and a longtime DataHub Champion's journey watching the project grow from a catalog into a context platform. Plus a first look at the new DataHub Metrics Catalog and what's coming at CONTEXT 2026. What's covered: Metrics Catalog Teaser → Shirshanka Das (DataHub CTO & Co-Founder) introduces the new DataHub Metrics Catalog, making business metrics first-class, governed entities with owners, lineage, and certified definitions ready for AI agents. Miro Spotlight → Ronald Angel (Miro) on taking natural-language-to-SQL accuracy from under 40% to over 90% by using DataHub as the context and routing layer, the 900-question eval bank, metadata signals, quality gates, and semantic views. JET Spotlight → Maksym Oliinyk (Just Eat Takeaway) on running DataHub self-hosted at scale as a production metadata service, governance-as-code, the operational gotchas of ingestion and resilience, and where their data product marketplace is headed. Community Spotlight → Tim Bossenmaier (DataHub Champion) on his journey from first discovering DataHub, to contributing back to the project, to watching it evolve from a catalog into a context platform. Key takeaways: Why text-to-SQL accuracy is a context and routing problem, not a code-generation one How metadata signals, semantic views, and quality gates lift agent accuracy in production What it takes to operate DataHub self-hosted at scale: ingestion, resilience, and lineage How governance-as-code keeps business metadata trustworthy and reviewable Why contributing to open source pays off, for you and your company A first look at the DataHub Metrics Catalog and CONTEXT 2026 Featured speakers: Ronald Angel — Miro (Lead Product Manager) Maksym Oliinyk — Just Eat Takeaway (Data Engineer, Data Platform Governance) Tim Bossenmaier — DataHub Champion (Data Architect, Cloudflight) Shirshanka Das — DataHub (CTO & Co-Founder) FIND MORE ABOUT DATAHUB 🚀 Join the DataHub open source community ▶︎ https://hubs.la/Q03YmhK20 🔍 Take the DataHub interactive product tour ▶︎ https://hubs.la/Q03Ymjqp0 ➡️ Explore DataHub's blog for top data and AI insights ▶︎ https://hubs.la/Q03YmjLN0 KEEP IN TOUCH! X: https://x.com/DataHubCloud LinkedIn: https://www.linkedin.com/company/datahub-cloud/ Chapters: 0:00 Welcome & Community Updates 01:15 Agent Hackathon & CONTEXT 2026 02:57 Join the Community 03:12 Metrics Catalog Teaser 05:43 Agenda 06:38 Miro Spotlight: Enhancing Text-to-SQL Accuracy (Ronald Angel) 19:06 JET Spotlight: Running DataHub at Scale (Maksym Oliinyk) 43:16 Community Spotlight: From Catalog to Context (Tim Bossenmaier) 54:38 Closing Remarks