You will design, build and run the AI agents at the heart of our work: assistants that answer, book and sell on WhatsApp, voice and web, and the automations behind them. You will take an agent from a first prompt to production, and keep it measurable, safe and useful once real people start talking to it.
What you will do.
Six things you will own from the start, in plain words. The scope grows with you.
- 01 Build agents that answer, book and sell across WhatsApp, voice and web.
- 02 Design prompts, tools and retrieval so agents act on real business data.
- 03 Write evaluations and monitoring, so every release is measurably better.
- 04 Decide with the team where an agent hands over to a person.
- 05 Turn what works into reusable parts of BLUbot and BPI.
- 06 Work with designers in Milano and engineers in Guangzhou when an agent meets a device.
What you bring.
Nobody matches a list perfectly. If you have most of this and proof you can learn the rest, apply.
- Solid software engineering, with production code you can show.
- Hands-on work with large language models: prompting, tool use, retrieval and evaluation.
- Good judgment about privacy, safety and the ways agents fail.
- Clear writing in English.
- Care for the person on the other side of the conversation.
- Voice or messaging platforms in production.
- Italian or Mandarin.
- Building with Claude.
Your first week.
How we hire.
Six steps, no surprises. A human reads every application, and if we say no after we meet, we tell you why.
Apply for
this role.
A person reads every application. Send something you made, not only a CV, and you get a clear yes or no.
Don’t see your role?
Tell us one thing you would build here, and send proof you can build it.