What runs today
Persistent Brain
Vector store, semantic recall, topic clusters — with context injection.
Multi-Host Lab
Distributed GPU inference across multiple machines. Always-on. GPU-free/batch separation.
Multi-LLM Router
Tier routing across frontier + local. REST API with auth, RBAC, rate limiting.
CI/CD & Resilience
Auto-deploy, schema drift gates, reproducible builds. Mechanics, not claims.
Updates
Rebrand abgeschlossen
2026-06-06