Rapida v0.1.0 — Foundations

This is the first public release of Rapida, an open-source, end-to-end Voice AI orchestration platform built for real production systems.

Rapida focuses on the hard problems teams face when deploying voice agents at scale: reliability, observability, and control.

What’s included in this release

1. Real-time voice orchestration
2. Streaming audio pipelines with low-latency handling
3. Built-in VAD and end-of-speech detection

Pluggable AI stack

1. Speech-to-Text (STT) integration
2. LLM-based agent reasoning
3. Text-to-Speech (TTS) streaming responses

Production-grade foundations

1. Stateful call handling and routing
2. gRPC-based internal APIs for performance and reliability
3. Designed for horizontal scaling

Observability by design

1. Visibility into audio flow, latency, and agent execution
2. Debuggable interactions instead of black-box behavior