C++ Machine Learning Engineer, Models Training

Tenstorrent; Warszawa; Hybrid; PLN 180,000 - 280,000; Full time

Benefits offered by this job

Highly competitive compensation

Benefits package

Equal opportunity employer

Job summary

A leading AI technology company is looking for a C++ Machine Learning Engineer to join their innovative team in Warsaw or Gdansk, Poland. This role involves extending and optimizing ML training frameworks and improving model performance on custom hardware. Candidates should be strong in C++ and have an interest in low-level systems programming. The position offers a hybrid work model and is open to remote candidates on a case-by-case basis.

Qualifications

• Strong in C++ and low-level systems programming, especially in performance-critical code.
• Comfortable with concepts in tensors, memory layout, and compiler graphs.
• Familiar with PyTorch and eager to understand framework hardware mapping.

Responsibilities

• Extend and optimize ML training framework with new operations and training features.
• Debug and tune model performance on Tenstorrent chips.
• Collaborate with compiler and kernel teams for model compilation and performance.
• Support integration of real‑world models into production.

Skills

Strong in C++

Low-level systems programming

Performance-critical code

Familiar with PyTorch

Problem-solving

Tools

PyTorch

Compiler tooling

Job description

Tenstorrent is leading the industry on cutting‑edge AI technology, revolutionizing performance expectations, ease of use, and cost efficiency. With AI redefining the computing paradigm, we develop a high performance RISC‑V CPU from scratch and aim to build the best AI platform possible.

As a C++ Machine Learning Engineer on the AI Models team, you’ll work on the training framework behind our most advanced models, writing high‑performance C++ code, shaping new layers and operators, and helping models scale across our custom silicon.

This role is hybrid, based out of Warsaw or Gdansk, Poland. We also consider remote candidates on a case‑by‑case basis.

We welcome candidates at various experience levels. During the interview process, candidates will be assessed for the appropriate level and offers will align with that level.

Who You Are

• Strong in C++ and low‑level systems programming, especially in performance‑critical code.
• Comfortable thinking in tensors, memory layout, and compiler graphs.
• Familiar with PyTorch and curious about how frameworks map to hardware.
• A builder who enjoys solving technical puzzles and digging into the details.

What We Need

• Extend and optimize our ML training framework with new ops, layers, and training features.
• Debug and tune model performance on Tenstorrent chips.
• Work with compiler and kernel teams to make sure models compile and run as expected.
• Support integration of real‑world models and help bring them into production.

What You Will Learn

• How ML frameworks and compilers connect at the system level.
• How to translate training workloads into low‑level operations optimized for custom silicon.
• How large‑scale model training works under the hood, from memory layout to operator fusion.
• What it takes to build infrastructure that supports fast iteration in research and production.

Tenstorrent offers a highly competitive compensation package and benefits, and we are an equal opportunity employer.

Tenstorrent is a U.S. export‑controlled technology company and employment may be contingent upon your citizenship or residency status. If any employment status makes this role impossible, any offer of employment will be rescinded.