Skip to main content
Track 01 — Reliability

Reliability engineering for regulated finance

Incident pipelines, SLOs and error budgets for regulatory processes, operational resilience under DORA and FCA regimes. Making critical systems boring, on the record.

reliability posts →
Track 02 — Applied AI

AI systems that hold up in production

Production LLM engineering: deterministic gates before advisory models, evals in CI, agents you can audit.

llm-council on github →
Open source — flagshipall projects →

llm-council

Multi-LLM deliberation with peer review and synthesis — library, MCP server, or HTTP API.

pythonmcpevals
github ↗

luminescent-cluster

Three-tiered memory architecture for persistent AI coding context — agents that stop forgetting.

pythonragmemory
github ↗

stentorosaur

Status monitoring as a Docusaurus plugin — uptime, incidents and notifications on GitHub Issues, zero infrastructure.

typescriptstatussre
github ↗
Recent writing — two tracksall posts →
2026-03-26[llm-ops]Memory as a Service: Multi-Agent CollaborationWhen AI agents need to share context, hand off tasks, and collaborate without stepping on each other — the architecture behind MaaS.2025-06-10[reliability]Beyond Uptime: Monitoring Business ProcessesUsers don’t care about your infrastructure. Reframing status monitoring around the business processes that actually matter.2024-05-21[reliability]Monitoring vs ObservabilityReactive monitoring tells you something broke. Observability tells you why. Where each fits in a mature reliability practice.

This site practices what it preaches

amiable.dev monitors itself every five minutes, files its own incidents as GitHub issues, and publishes its uptime history — the same patterns I write about, running in production here.

view live status