TySVA - TypeScript Voice Assistant🪄

What is TySVA - TypeScript Voice Assistant🪄?

TySVA is a voice- and text-enabled assistant that helps users learn TypeScript through natural conversation, leveraging AI agent workflows. It runs locally via Docker or source code, integrating Qdrant, LinkUp, ElevenLabs, LlamaIndex, and MCP servers.

How to use TySVA - TypeScript Voice Assistant🪄?

Clone the repository, set up a .env file with Groq, ElevenLabs, and LinkUp API keys, then launch using the provided Docker compose or conda scripts. The application becomes accessible at http://localhost:7999/app.

Key features of TySVA - TypeScript Voice Assistant🪄

• Voice input and output via ElevenLabs
• Textual input and output option
• Local Qdrant vector database with full TypeScript documentation
• Web deep search via LinkUp
• Agent workflow orchestrated with LlamaIndex
• MCP servers for vector and web search automation

Use cases of TySVA - TypeScript Voice Assistant🪄

• Learn TypeScript by asking natural language questions
• Retrieve relevant documentation from a local vector database
• Search the web for current TypeScript topics and examples
• Practice TypeScript concepts using voice commands

FAQ from TySVA - TypeScript Voice Assistant🪄

What APIs are required to run TySVA?

You need API keys from Groq, ElevenLabs, and LinkUp, which must be placed in a .env file.

Can I use TySVA without voice input?

Yes, TySVA supports both voice and textual input/output; you can use it solely through text.

How is the application deployed?

It runs locally using Docker containers or source code with conda. After setup, it is accessible at http://localhost:7999/app.

What license does TySVA use?

The software is provided under the MIT license.

Does TySVA use MCP servers?

Yes, it uses MCP servers to automate vector search (Qdrant) and web search (LinkUp) within the agent workflow.