Careers
Build the voice of the other half.
Most speech models are good in about ten languages. We build for the rest — 80+ languages, hundreds of voices, in a model people already run in production. We are a small team in Bengaluru, and we are hiring.
How we work
- Ship it
- A model that is not in someone's hands is a claim, not a result.
- Listen to the output
- Metrics rank checkpoints. Ears decide which one ships.
- Go where the data isn't
- The languages everyone already serves are not the interesting problem.
- Small team, wide scope
- You will own more than your title, and less will be decided for you.
- Write it down
- A written argument survives the meeting. Most of ours start that way.
- Say it plainly
- Get to the point, and assume the other person meant well.
01.
02.
03.
04.
05.
06.
What we offer
- Competitive salary, discussed early in the process.
- Direct ownership of a system in production, not a slice of a backlog.
- A small team and a short path from idea to shipped.
- Hardware and compute you need to do the work.
- Based in our Bengaluru office.