Embedded AI Engineer Intern
Port and optimise neural network models to run onboard our spacecraft computing platforms, under hard compute, memory, and power limits.
About the position
We are looking for an Embedded AI Engineer Intern to help run modern AI workloads directly onboard our spacecraft computing platforms. Inference in orbit has to fit inside hard limits on compute, memory, and power, hold up under real-time constraints, and stay accurate enough to be worth flying. Making that trade-off work is the core of this role.
In this role, you will port and optimise neural network models for constrained embedded targets: quantization, pruning, operator fusion, and accelerator-specific compilation. You will profile and benchmark inference performance on flight-representative hardware, build the tooling and pipelines that make deployment repeatable, and help define how model performance is measured and validated before it goes onboard.
This internship offers the opportunity to work directly with our software, hardware, and avionics engineering teams and to gain hands-on experience with edge AI deployment, model optimisation, and the validation activities used in advanced aerospace applications.
Required skills
- Final-year or gap-year student in computer science, electrical engineering, embedded systems, machine learning, or related field
- Good understanding of deep learning fundamentals and common model architectures (CNNs, transformers, and similar)
- Experience with ML frameworks (PyTorch, TensorFlow, ONNX, or similar)
- Comfortable with Python and C/C++, and working in a Linux environment
- Experience deploying models on embedded or accelerated hardware (NVIDIA Jetson, embedded GPUs, FPGAs, or similar) is a plus
- Familiarity with model optimisation and inference stacks (quantization, pruning, TensorRT, TFLite, or similar) is a plus
- Ability to profile and reason about performance: latency, throughput, memory footprint, power
- Knowledge of space or radiation constraints on computing is a plus, but not required
- Curious, methodical, and eager to work on multidisciplinary hardware-software projects
Why join Coros Space
Joining Coros Space means joining a team designing high-performance computing systems for tomorrow's satellites. As an intern, you will:
- Work closely with experienced engineers in avionics, software, and hardware
- Contribute directly to a flight-representative onboard computing stack
- Learn how real spacecraft systems are designed, tested, and validated
- Have meaningful impact in a startup environment where your work matters
How to apply
Send your CV to careers@coros-space.com.
The subject line of your email must be exactly:
[COROS-INT26-EAI] Embedded AI Engineer - Intern 26
Applications sent without a CV attached, or without the position title and code in the subject line, may not be processed correctly.