This project provides an open-source platform for building and managing advanced, goal-oriented AI agents for both voice and chat. It targets developers and businesses looking to create sophisticated conversational workflows beyond simple Q&A, enabling complex task execution and system integration. The platform offers a visual workflow canvas, RAG knowledge base capabilities, and native telephony integration.

How It Works

Intervo.ai utilizes a node-based visual editor to design complex conversational flows, allowing for routing based on user intent and orchestrating specialized "sub-agents." It supports goal-oriented dialogues, ensuring agents complete specific tasks. The platform integrates with various speech and text-to-speech services, LLM providers via LangChain, and includes a RAG knowledge base powered by vector search with ChromaDB for training agents on custom data.

Quick Start & Requirements

• Installation: Docker Compose is the recommended method.
• Prerequisites: Docker, Docker Compose, Git, FFmpeg.
• Configuration: Requires setting environment variables, including API keys for AI providers (OpenAI, Groq, Gemini, Anthropic), speech services (Deepgram, AssemblyAI, Google STT, AWS Polly, Microsoft Speech, ElevenLabs), Twilio for telephony, Stripe for billing, and object storage.
• Access: Frontend at http://localhost:3000, Backend API at http://localhost:3001, RAG API at http://localhost:4003.
• Documentation: https://docs.intervo.ai/

Highlighted Details

• Multimodal AI agents supporting real-time voice calls and text-based web chat.
• Flexible LLM integration with support for OpenAI, Groq, Google Gemini, and Anthropic.
• Native telephony integration with Twilio for inbound/outbound calls.
• Embeddable, customizable React-based web widget for website integration.

Maintenance & Community

• Community: Discord community available.
• Contributing: Contributions are welcomed; a contributing guide is provided.
• Core Contributors: Manjunath M, Hakhil Nizeem, Rahul, Amar, Geethu Sebastian, Vasanth.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license generally permits commercial use and linking with closed-source projects.

Limitations & Caveats

The project is actively under development with a roadmap indicating ongoing work on documentation, setup simplification, and test coverage (Phase 1). Features like agentic tools, WebRTC integration, and mobile SDKs are planned for future phases.