Voice AI & Speech
Speech recognition, text to speech and real-time voice agents, including Indian-language speech.
Beginner
Start here. No background needed beyond school maths.
Azure AI Speech: Overview
What a managed speech service offers: speech to text, text to speech, translation and speaker recognition, and how to call them from an app. Includes quickstarts.
Sign up to see the source
Hugging Face Audio Course
A free course on working with audio in machine learning: how sound becomes data, then speech recognition, text-to-speech and audio classification with transformers.
Sign up to see the source
Advanced
Once the basics are comfortable.
Common Voice
A free, crowd-sourced speech dataset in many languages, including Hindi and Telugu, that you can download to train or test speech recognition.
Sign up to see the source
LiveKit Agents Documentation
Build a voice agent that listens, thinks with a language model and speaks back in real time over WebRTC, with turn detection and tool calling.
Sign up to see the source
Pipecat Documentation
An open-source framework for real-time voice and multimodal agents: chain speech recognition, a language model and speech synthesis into one low-latency pipeline.
Sign up to see the source
Whisper: Robust Speech Recognition
••••'s open-source speech recognition model that transcribes and translates many languages, including Indian ones. The repository explains the model sizes and how to run it.
Sign up to see the source
Expert
Research-level depth and specialist material.