Job Summary
Role Description
We are looking for experienced C++ / PyTorch AI Framework Engineers to develop, debug and validate components of an AI software stack targeting hardware accelerators.
The engineer will work on framework implementation, defect investigation, automated testing and software validation across C++ and Python components.
Key Responsibilities
-
Develop features and enhancements for PyTorch/AI framework components.
-
Implement and maintain C++ and Python software.
-
Analyze and resolve complex software defects.
-
Implement and debug deep-learning operators and related functionality.
-
Develop unit and integration tests.
-
Perform framework/software validation.
-
Debug software on dedicated hardware/test platforms where required.
-
Analyze regressions and verify functional and numerical correctness.
-
Prepare code patches and technical analysis.
-
Identify cross-component dependencies
- Participate in code reviews, planning and engineering discussions.
Skill Requirements
-
Strong C++ programming skills.
-
Good Python skills.
-
Linux development experience.
-
Understanding of deep-learning frameworks and concepts.
-
Strong debugging/root-cause analysis.
-
Unit and integration testing.
-
Git/source-control experience.
-
Ability to independently deliver engineering tasks.
-
Good English communication skills.
-
Ability to work onsite in Gdańsk when hardware/laboratory access is required.
Other Requirements
-
PyTorch development/framework experience.
-
Tensor/operator development.
-
GPU/accelerator programming.
-
CUDA, SYCL, oneAPI, ROCm or OpenCL.
-
Intel GPU / Gaudi / Habana exposure.
-
HPC / numerical computing.
-
CMake / GDB / pytest.
-
Numerical correctness and floating-point concepts.
-
Open-source/upstream development experience.