Launched this week
AVA - AI Voice Agent for Asterisk
Open-source AI voice agent for any phone system
35 followers
Open-source AI voice agent for any phone system
35 followers
AVA is a self-hosted, open-source AI voice agent that plugs directly into your Asterisk or FreePBX system. No cloud telephony migration required. Choose from 6 production-ready configurations: OpenAI Realtime, Deepgram, Google Gemini Live, ElevenLabs, Local Hybrid, or 100% fully Local. Modular STT/LLM/TTS pipeline — swap providers without changing code. Try it live: dial (925) 736-6718.
@hkjarral finally someone is building for asterisk directly. trying to bridge modern llms with legacy pbx systems usually feels like a hack. having full control over the stack without a middleman is a huge win for us. Love it
how's the jitter buffer handling the llm stream? if you've solved real-time on asterisk, that's impressive. great release. @hkjarral
@vikramp7470 All settings are exposed via UI and tunable, But for baseline, LLM-driven TTS audio arrives in unpredictable bursts. The jitter buffer absorbs that burstiness and drains it at a steady 20ms rhythm into Asterisk — so the caller hears smooth, continuous audio.
It waits for ~120ms of audio to accumulate before starting playback (prevents initial stutter), auto-tunes its size based on each provider's streaming pattern, and fills gaps with silence if the provider falls behind temporarily.
@hkjarral really smart approach to the burstiness.
User of AVA for ISP Triage, works great! Strongly encourage use of this system. I use it with integrations into Ollama vs the known AI systems, and interfacing to help identify whether or not a connection is at its quality limit, create tickets, and check billing status. This is done by knowing the phone number, and such aligning with the billing system.