6 projects
ridgepoint
LLM inference sizing that models the engine, not just the weights: engine-aware VRAM capacity, KV cache (GQA/MLA), and roofline intervals — calibrated to ~1% against real vLLM.
prefixcash
Measure, diagnose and fix your LLM prompt-cache hit rate — and see the dollars you save.
diva-ai
Diva SDK for Python (thin client) — build agents on the Diva platform over a hosted gateway (bearer token, no local engine)
diva-ai-sdk
Reserved by the Diva platform team for the official Diva AI SDK (agent harness: tools, RAG knowledge bases, frame-flows). Official release coming soon.
diva-sdk
Reserved by the Diva platform team for the official Diva AI SDK (agent harness: tools, RAG knowledge bases, frame-flows). Official release coming soon.
explicit-decision-protocol
Reference Python SDK for the Explicit Decision Protocol (EDP)
