Coming soon · Applied AI roadmap
EvalGate
Status: not built yet
Objective
Build a reproducible evaluation platform for LLM/agent systems.
Proof sought
Versioned datasets, graders, configuration comparisons, quality/cost/latency, a CI quality gate.
Planned stack
PythonFastAPIPydanticPostgreSQL / pgvectorNext.js / TypeScriptDockerGitHub Actions
Shared stack across the Applied AI roadmap (not yet implemented on this specific project).