Software Engineer - Autonomy, MLOps - #1777977

About Delian

We live under a new world order. To protect Europe and its allies we're building a new defense prime. One tailored for the era of autonomous warfare.

While building for the future, our culture emulates the defense industry's roots: mass scale; operational excellence; execution speed; rapid technology adoption.

To strive for this we combine expertise in autonomy, electronic warfare and rapid manufacturing. Paired with young, passionate and mission driven engineering and technical talent.

We have offices in London & Athens.

About The Role

We are seeking a Software Engineer to join our Autonomy team with a focus on real-time detection systems. In this role, you’ll develop core components of the perception stack that allow our autonomous systems to detect, classify, and track objects across varied and challenging environments. You’ll work closely with teams across perception and sensor integration to deliver reliable detection pipelines for deployment in the real world.

This role is ideal for engineers who enjoy solving vision and sensor-fusion challenges at the core of autonomous perception.

In this role, you will get to

• Develop detection pipelines for real-time inference using sensor modalities such as cameras and radars.
• Integrate and optimize deep learning models (e.g., 2D/3D object detection, segmentation) for deployment on embedded platforms.
• Build robust, scalable perception systems capable of operating under variable lighting, weather, and scene complexity.
• Collaborate with tracking, planning, and calibration teams to ensure consistent handoff of perceptual data throughout the autonomy stack.
• Contribute to tools and infrastructure for performance analysis, dataset creation, labeling, and model evaluation.
  
  

Must have experience:

• Strong programming skills in Python and C++, especially for high-performance robotics or real-time applications.
• Experience implementing or integrating object detection and/or semantic segmentation models in production environments.
• Familiarity with machine learning frameworks such as PyTorch or TensorFlow, and knowledge of model training/inference workflows.
• Proficiency with NVIDIA’s edge computing stack, including CUDA, TensorRT, DeepStream SDK for real-time multi-stream video analytics and pipeline integration.
• Understanding of 3D geometry and computer vision fundamentals (e.g., projections, point cloud processing, depth estimation).
• Experience working with multi-modal sensor data (camera, LiDAR, radar) in perception pipelines.
• Ability to diagnose system-level failures and improve detection reliability across varied operational conditions.
  
  

Nice To Have Experience

• Experience deploying and optimizing perception models on NVIDIA Jetson (e.g., Xavier, Orin) or similar embedded platforms.
• Familiarity with ROS/ROS2 and integrating detection nodes into larger autonomy frameworks.
• Understanding of sensor calibration, data association, and perception uncertainty modeling.
• Experience with active learning, dataset bootstrapping, or labeling tools and pipelines.
• Contributions to open-source projects in robotics, computer vision, or ML infrastructure.