Blog
Engineering in public.
Deep dives on building Logic64 — architecture decisions, infrastructure trade-offs, and the reasoning behind governance-first AI generation.
First posts shipping in June 2026.
Coming up
EngineeringComing soon
Why we built a governance layer before writing a single line of generation code
The AI hallucination problem in code generation is not a model problem — it's an architecture problem. Here's how we approached it.
ProductComing soon
Process isolation as a first-class feature: how Logic64 keeps generations stateless
Every workspace generation in Logic64 runs in an isolated subprocess. Not for performance — for correctness. Here's why that matters.
EngineeringComing soon
BullMQ at the core: designing a job queue that survives restarts
SSE streaming, job persistence, and graceful failure — the infrastructure behind Logic64's generation pipeline.
Get notified when we publish.
No marketing. Just engineering posts when they drop.