From Pain to Performance: Solving the 6 Pitfalls of Cloud BI Migration

Moving BI to the cloud shouldn’t be “same dashboards, new bill”. In this session, Martin Mahler (CEO & Founder, Astrato Analytics) and Alejandro Martinez (CEO & Founder, Propelling Tech) unpack why lift-and-shift fails, where costs hide, how performance and UX often degrade, and the plays that actually deliver adoption: live/ELT pushdown, a governed semantic layer, warehouse-enforced security (RLS/CLS), and app-like experiences. You’ll learn how to: – Avoid lift-and-shift traps and redesign for cloud-native BI – Control TCO with the right unit metrics – Kill “multiple versions of truth” with a semantic layer + lineage – Engineer for concurrency and speed (MVs, partitions, caching) – Put governance where data lives (IdP SSO, RLS/CLS in the warehouse) – Keep your stack open and AI/embedding-ready 🔖 Chapters 00:00 Welcome & speaker intros 02:18 Why cloud BI now (vendor push + AI readiness) 06:14 Cloud-native vs lift-and-shift (re-design needed) 09:58 Performance realities in the cloud 12:49 Cost reality (licenses, egress, AI add-ons; performance = cost) 19:44 Modernize during migration; what to migrate vs retire 26:00 Shift-left: semantic layer, medallion, self-service that works 33:26 Product-like UX: writeback, workflows, adoption metrics 35:31 Governance in the warehouse (SSO, RLS/CLS, zero-copy) 42:01 Future-proofing: open architecture, AI/BI, data apps, vendor roadmaps 49:17 Six-pitfall recap & action checklist 51:02 Closing & next steps