[rokoss21]

I’m working on a deterministic execution layer for AI systems.

The goal is to treat LLMs as constrained components inside explicitly defined workflows: strict input/output schemas, validated DAGs, replayable execution, and observable failure modes. Most instability I’ve seen in production AI systems isn’t model-related — it comes from ambiguous structure around the model.

We’re exploring this through a project called FACET, focused on making AI behavior testable and debuggable rather than probabilistic and opaque.

Early days, but the direction is clear: less magic, more contracts.