AI Developer
We're looking for an AI Engineer to build and own our real-time speech-to-text system from the ground up. Our audio comes from military radio traffic in the field: noisy, degraded, and unlike anything a standard speech model is trained on. You'll be responsible for adapting and fine-tuning models to handle that reality, and for getting them running reliably on resource-constrained, on-premise hardware — not just in a notebook, but in production, in the field. This is a hands-on, end-to-end role. You'll define how success is measured, build the pipeline from raw recordings to a deployed model, and iterate based on real performance data. You'll work closely with a small team in a fast-moving, seed-stage environment where ownership matters more than process.
Details
- Location: Paris, France · On-site
- Employment: Full-time
- Seniority: Mid-Senior Level
- Visa & work authorization: Relocation support available.
A seed-stage European defence startup building modern software for hostile environments (military and police units). The company is currently operating in stealth and does not work on weapons systems.
What we're looking for
- Professional experience programming in Python
- Experience implementing Voice AI or speech-to-text systems, including fine-tuning models rather than wrapping an off-the-shelf API
- A track record of owning at least one ML or data-heavy project end-to-end, from defining evaluation metrics through to shipped results
- Willingness to work on-site in Paris
Nice to have
- Experience deploying models in resource-constrained or on-premise environments
- Background in audio processing, voice activity detection, or diarisation
- French language proficiency