Giulia runs as a persistent OTP daemon — AST indexes, knowledge graph, and embeddings stay warm in ETS between sessions. No cold starts, no re-indexing per query.

Core: two-node BEAM cluster (worker + monitor), three-tier storage (ETS → CubDB → ArcadeDB), Sourceror for AST parsing, pluggable LLM providers, 78 self-describing REST endpoints.

Beyond static analysis, it connects to any running BEAM node via distributed Erlang and fuses live process data with the static graph — which hot process, which module, what breaks if you touch it.

Apache 2.0.

https://github.com/thatsme/giulia