Skip to content
tarıtas
Careers

Build voice AI that ships.

Fifteen senior engineers in Central India, working NA, EU, and AU overlap hours. What you build here answers phones for a city of 230,000 by week eight, not a backlog. Small team, production from the first sprint, one engineering lead per engagement.

How to apply
For freshers

What we expect before you apply.

No years-of-experience theater. Six things, all learnable in a few focused weeks, all free.

01 Fundamentals that survive contact with latency Data structures, networking basics, what a websocket is. Voice runs on an 800ms budget; you should know where milliseconds go.
02 One language, properly Python or TypeScript, your pick. Comfortable reading other people's code, not just writing your own.
03 Something real built with an LLM API End to end, not a tutorial clone. Bring the repo; we will read it before we read your resume.
04 The Anthropic courses, completed Anthropic Academy is free and certificates are issued. Finish the API course at minimum; it is the floor we interview from.
05 The LiveKit quickstart, built and understood A working voice agent takes under an hour. Being able to explain what each piece does takes longer; that is the part we ask about.
06 Writing we can follow A README that explains what, why, and how to run it. Production voice AI is documented or it does not pass a security review.
Prepare

The four things to finish before applying.

This is the exact preparation list we would give a friend. Total cost: zero. Total time: two to three focused weeks.

Anthropic Academy · free · certificate

Claude with the Anthropic API

The full API lifecycle: first call to streaming, tool use, and production patterns. The single most useful course for our work.

Open → Anthropic Academy · free · certificate

Introduction to Model Context Protocol

How Claude connects to external tools and data. Our agents act through client systems; MCP is how that thinking is structured.

Open → LiveKit docs · free

Voice AI quickstart

A working voice agent you can talk to in under an hour: STT, LLM, TTS, turn detection. Build it, break it, explain it.

Open → LiveKit · GitHub

LiveKit Agents framework and examples

The framework we run in production. Read the examples directory; the restaurant agent example is the closest to our daily work.

Open →
How to apply

One email. No forms, no portals.

enquiries@taritas.com

Subject line: “Application - Voice AI Engineer” and whether you are applying as a fresher or experienced. We read every application. A repo we can run beats a cover letter we have to read.

Attach or link
  • Resume, one page, PDF
  • Links to 1 or 2 repos you actually built
  • Course certificates, if you have them
  • Five lines on the most interesting thing you made