Ollaya launches open-source local runtime for Jev-style decision models with millisecond responses and TypeSafe API compatibility

Ollaya lets developers run decision models locally to ask typed questions about text or JSON and receive calibrated answers in milliseconds. The open-source tool runs on private hardware, offers Ollama-like model management, and supports TypeSafe-compatible endpoints and SDK usage.

The project reflects growing demand for lightweight, local AI components that handle classification and policy decisions without full token generation. Its API compatibility could make it easier for teams to shift latency-sensitive or privacy-sensitive workloads from hosted services to on-premise infrastructure.