Insights

Practical notes on AI agents, sovereignty, and production delivery for EU teams.

• RAG vs GraphRAG: choosing for your documents

  2026-10-15 — RAG or GraphRAG? Choose by question type, source quality, access rights and lifecycle cost—not architecture hype. A guide for CTOs and managers.

• How to evaluate an AI agent before launch

  2026-10-15 — Test business outcomes, tool use, permissions, cost and reliability before launching an AI agent. A practical guide for CTOs and managers.

• Human-in-the-loop for AI agents: when to ask for approval

  2026-10-14 — When should an AI agent ask for approval? A practical risk matrix for CTOs and managers, with exact-action gates and fewer unnecessary prompts.

• AI agent production readiness checklist: 24 checks before go-live

  2026-10-13 — 24 checks across data, security, evaluation, ownership, cost and adoption to decide if an AI agent is ready for production, with a go/no-go rule.

• AI agents in production: from POC to a system you run

  2026-10-12 — What changes between an AI agent demo and a production system: data, security, evaluation, ownership, cost and adoption. A practical guide.

• Giving AI agents access to your email: 7 questions to ask your AI expert

  2026-10-12 — Before an AI agent reads your mailbox, ask these 7 questions. Lessons from connecting Gmail and Infomaniak through MCP in a read-only pilot.