ML Product Engineer

The Challenge: Turn Research into Industrial IP Sigma Nova’s growth depends on our ability to translate research into Capabilities: reusable technical building blocks (pipelines, frameworks, interpretability tools) that become permanent assets for the company and our clients. Your mission: Transform research experiments into installed, documented, and monetizable capabilities that are ready for deployment, scaling, and client integration. What You’ll Do • Building the Capability Library: • Support the development of modular, reusable components emerging from research (e.g., EEG preprocessing pipelines, fMRI interpretability tools).

• Act as the guarantor of versionability, documentation, and reproducibility, ensuring that research outputs can be reliably reused and extended.

• Finetuning & Client Adaptation: • Execute and optimise finetuning pipelines (PyTorch, Hugging Face) for diverse domains.

• Adapt foundation models to specific client needs while maintaining performance and scalability.

• Backend & Inference: • Design high-performance inference servers (FastAPI, gRPC) and SDKs to expose capabilities seamlessly.

• Optimise for low latency, scalability, and ease of integration.

• Deployment & Ops: • Deploy models via Docker/Kubernetes on Scaleway to ensure a frictionless “Lab to Production” transition.

• Implement monitoring, logging, and maintenance for long-term reliability.