Making Your Data Ready for Agentic AI: trust first, then semantics, then writes
bibryam · x · 2026-10-16
Thoughtworks' Pramod Sadalage and Prem Chandrasekaran argue on martinfowler.com that data systems built for human analysts—who supply context and skepticism—don't fit autonomous agents that act confidently on whatever they're handed.
Key points
- AI readiness is capped by the weakest foundational data layer; fix trust before semantics.
- Proposes a layered architecture: a trusted data foundation, a context layer making shared vocabulary executable (domain ontology, semantic metrics compiled to SQL, a capability model with governed reads and audited actions), and an access layer that moves from read-only to writes only as governance matures.
- Requires continuous observability for governance and an auditable trace of agent-driven decisions.
More from coding & agent
- Connected steps are not a workflow: nobody designs the handoff — usehive · 2026-10-16
- Making Blender animations with Opus, Astra and Argon: how you prompt still matters — chongdashu · 2026-10-16
- Running PolyStrata + Cline coding agent on a $0-fancy laptop at ~90 tok/s — rpwoerk · 2026-10-16
- How do you attribute agent gains: model training or infrastructure? — jrxna · 2026-10-16
- Feed Claude your books, blogs and top engineers' code for claude.md — 'it now writes better code than me' — blaizedsouza · 2026-10-16
- Open-source /label plugin auto-names and colors Claude Code sessions in one step — hermanschutte · 2026-10-16