Jekyll2026-07-12T00:19:43+00:00https://productionnotes.dev/feed.xmlProduction NotesProduction engineering notes on enterprise AI agents, workflows, and SRE — by Sabith K S, Staff Engineer at StackGen.Sabith K SHow to Debug Multi-Stage AI Agent Workflows — Bring Up Like Hardware2026-07-10T17:00:00+00:002026-07-10T17:00:00+00:00https://productionnotes.dev/blog/bring-up-agent-workflows-like-hardwareSabith K SMaintaining Tokenomics with Aiden — Context Budgets as an Operating Model2026-07-09T17:00:00+00:002026-07-09T17:00:00+00:00https://productionnotes.dev/blog/maintaining-tokenomics-with-aidenSabith K SProve, Then Narrate — Deterministic Orchestration Over Autonomous Agents2026-07-08T17:00:00+00:002026-07-08T17:00:00+00:00https://productionnotes.dev/blog/evidence-gated-multiplane-rcaSabith K SEvidence-Based Verification — Don’t Trust Self-Report, Check the System2026-07-08T17:00:00+00:002026-07-08T17:00:00+00:00https://productionnotes.dev/blog/evidence-based-verificationSabith K SAI Incident Triage for SREs — What Actually Helps On-Call2026-07-07T17:00:00+00:002026-07-07T17:00:00+00:00https://productionnotes.dev/blog/ai-incident-triage-sreSabith K SYou Vibe-Coded the AWS Calls — Do You Know What IAM Permissions You Actually Need?2026-07-05T18:00:00+00:002026-07-05T18:00:00+00:00https://productionnotes.dev/blog/cce-cloud-entitlementsSabith K SWhy One JSON Repair Pass Isn’t Enough for Production Agent Tool Calls2026-07-03T17:00:00+00:002026-07-03T17:00:00+00:00https://productionnotes.dev/blog/json-repair-layersSabith K SContributing Back While Building a Commercial Product2026-07-01T17:00:00+00:002026-07-01T17:00:00+00:00https://productionnotes.dev/blog/open-source-ecosystemSabith K SWhy We Split Our Agent Runtime From Our Platform2026-06-30T17:00:00+00:002026-06-30T17:00:00+00:00https://productionnotes.dev/blog/aiden-platformSabith K STerraform for Agent Configuration — Infrastructure as Code Meets AI Governance2026-06-29T17:00:00+00:002026-06-29T17:00:00+00:00https://productionnotes.dev/blog/terraform-configSabith K SYou Can’t Debug What You Can’t See — Observability for AI Agents2026-06-28T17:00:00+00:002026-06-28T17:00:00+00:00https://productionnotes.dev/blog/observabilitySabith K SYour Agent Has Root — Defense-in-Depth for AI Agents That Wield Real Tools2026-06-27T17:00:00+00:002026-06-27T17:00:00+00:00https://productionnotes.dev/blog/defense-in-depthSabith K SThe HITL Paradox — When Human Approval Makes Agents Worse2026-06-26T17:00:00+00:002026-06-26T17:00:00+00:00https://productionnotes.dev/blog/hitl-paradoxSabith K STeaching Agents to Learn Without Fine-Tuning2026-06-25T17:00:00+00:002026-06-25T17:00:00+00:00https://productionnotes.dev/blog/skill-distillationSabith K SPensieve — Memory Management for AI Agents That Actually Forget2026-06-24T17:00:00+00:002026-06-24T17:00:00+00:00https://productionnotes.dev/blog/pensieve-memorySabith K SImplementing ReAcTree — 6 Production Bugs the Paper Didn’t Warn You About2026-06-23T17:00:00+00:002026-06-23T17:00:00+00:00https://productionnotes.dev/blog/reactree-bugsSabith K SArchitecture at Speed Without Drowning2026-06-22T17:00:00+00:002026-06-22T17:00:00+00:00https://productionnotes.dev/blog/anatomy-of-a-platformSabith K STOML Over YAML and PKL — How We Stopped Fighting Config and Started Shipping2026-06-21T17:00:00+00:002026-06-21T17:00:00+00:00https://productionnotes.dev/blog/toml-over-yamlSabith K SGo vs Python for AI Agents — Why We Chose Go2026-06-20T17:00:00+00:002026-06-20T17:00:00+00:00https://productionnotes.dev/blog/why-goSabith K S