Hey Product Hunt, I am the co -founder of Arjun, video.
I am passionate to launch my Open Source AI Voice Agent SD.
Today, the sound is becoming a new UI. We expect the agents to understand us, respond immediately, and work without interruption in the web, mobile and even telephone. But, to achieve this, develoop, developers have to sew together: STT, LLM, TTS, HTTP’s closing points and a prayer.
As a result, often the result of agents that robbed, deceive and fail in the product environment without observation.
So we made something to solve it: Infrastructure from the end to the end for the construction, deployment and monitoring of your AI Voice Agents
This is what it offers:
Global WebRTC Infra with <80ms Litanisi
The ancestral turning detection, vAD, and noise pressed
Modular Pipelines for STT, LLM, TS, Avatar, and Real Time Model Switching
Bullet In Chopped + Memory for Grounding and False Resistance
SD for Web, Mobile, Alliance, IOT, and Telephone – Glow Code is not required
Agent Cloud Self with one-click deployment deployment with unlimited scale-or full control
Think about it as walking from walkie talkie to modern cell towers that handle thousands of calls.
Videos D provides you with infrastructure for building acoustic agents that actually work in the real world.
I would like your thoughts and questions! Glad to make a deep dive in architecture, use matters, or crazy edge matters with which you are struggling.