https://pedropizarrodiaz.it/blog/ Practical field notes on taking AI agents from demo to governed production. en Sat, 29 Aug 2026 12:00:00 +0200 https://pedropizarrodiaz.it/blog/ai-agent-production-raci.html https://pedropizarrodiaz.it/blog/ai-agent-production-raci.html Sat, 29 Aug 2026 12:00:00 +0200 Five production decisions that need explicit ownership: risk, evaluation, access, incidents and adoption. https://pedropizarrodiaz.it/blog/ai-agent-metrics-before-production.html https://pedropizarrodiaz.it/blog/ai-agent-metrics-before-production.html Sat, 29 Aug 2026 11:00:00 +0200 A three-level model for technical reliability, task quality, operating value and safety. https://pedropizarrodiaz.it/blog/ai-agent-test-gaming-claim-coverage.html https://pedropizarrodiaz.it/blog/ai-agent-test-gaming-claim-coverage.html Sat, 29 Aug 2026 10:00:00 +0200 Four ways green tests can fail to prove the behaviour a release decision depends on. https://pedropizarrodiaz.it/blog/agent-communication-contracts.html https://pedropizarrodiaz.it/blog/agent-communication-contracts.html Sat, 29 Aug 2026 09:00:00 +0200 A deterministic boundary for handoff, status, decision and debug messages between agents. https://pedropizarrodiaz.it/blog/why-ai-pilots-fail-after-deployment.html https://pedropizarrodiaz.it/blog/why-ai-pilots-fail-after-deployment.html Fri, 28 Aug 2026 12:00:00 +0200 The model can work while the operating change fails.