Machine Learning Engineer (Computer Vision) - Real-Time Defence AI

ML/Computer Vision, Deep Learning Frameworks, Image/Video datasets, MLOps

Paris (Hybrid)

50,000 – 70,000 €

We are partnering with a highly innovative deep-tech startup, founded in 2024, that specializes in real-time, deployable AI solutions for critical defense operations (Intelligence, Surveillance, Reconnaissance - ISR, and Counter-UAS systems). Having successfully raised €2M in pre-seed funding, they are scaling their engineering team to meet increasing demand from NATO-aligned military clients.This is a chance to join a mission-driven company where your work moves quickly from the lab to live field exercises and deployment.

🎯 The Role & Mission

As a Machine Learning Engineer specialising in Computer Vision, you will be responsible for the entire ML lifecycle across high-stakes product lines, focusing on deployment to edge compute units (Jetson, RPi) and cloud GPUs.

You'll be the key driver in ensuring their deep learning models—covering detection, segmentation, and tracking—are optimised for real-time performance and maximum reliability in operational conditions.Key Responsibilities Include:Training & Pipeline Ownership: Improve and maintain robust training pipelines for vision models and define automated evaluation frameworks.

• Edge Optimisation: Aggressively optimise models for inference on embedded hardware using techniques like quantisation, pruning, and weight conversion.
• Data Governance: Drive best practices for dataset versioning, triage, synthetic data generation, and annotation workflows.
• Deployment & MLOps: Contribute to full system integration, embedded workflows, and implement monitoring/feedback loops for field reliability.
• Cross-Functional Collaboration: Work hand-in-hand with Software, Hardware, and Field Operations teams to achieve rapid iteration from development to live deployment.

 🌟 Candidate Profile:

The Ideal Fit

We are looking for a highly pragmatic, field-oriented engineer who is ready to take ownership of the full ML pipeline and deliver production-level code, not just research prototypes

.Must-Have Skills:

• Minimum of 2+ years experience as an ML Engineer or Computer Vision Engineer, with a proven track record of shipping models into real usage.
• Expertise in Python and deep learning frameworks (PyTorch preferred).
• Hands-on experience training, optimising, and evaluating models (detection, segmentation, multi-object tracking) on large-scale image/video datasets.
• Strong foundation in MLOps fundamentals: Docker, Git best practices, CI/CD, and experience with experiment tracking/data governance tools (e.g., ClearML, MLflow, W&B, DVC, or equivalent). Highly Desirable Experience:Deployment experience on edge computing hardware (e.g., NVIDIA Jetson / RPi).
• Familiarity with optimization toolkits like TensorRT / ONNX Runtime and quantization workflows.
• Understanding of real-time performance constraints (latency budgets, thermal/power envelopes) where reliability is paramount over benchmark scores.

• What's Offered:
• Direct exposure to impactful, real-world ISR and Counter-UAS projects.
• End-to-end involvement across the ML lifecycle, from data prep to deployment on embedded systems.
• A collaborative, multidisciplinary environment with experts in AI, defence, and robotics. 

Interested in applying your computer vision expertise to mission-critical, real-time systems? Please apply.