Nikita Shulga

Open-source work

Shulga is one of PyTorch's core maintainers, and the project's governance documentation also names Shulga as a maintainer for two areas: the Apple M1/MPS/Metal backend and Build + CI. The repository's CODEOWNERS file assigns Shulga ownership of the MPS and Metal source directories.¹⁵

A 2025 PyTorch event page described Shulga as a software engineer at Meta responsible, among other things, for PyTorch releases and continuous integration, and as a core maintainer.³

Releases and community

Shulga has presented PyTorch releases to the community in live question-and-answer sessions, including PyTorch 2.7 in April 2025, PyTorch 2.10 in January 2026 and PyTorch 2.11 in March 2026. For the PyTorch 2.14 session in September 2026, which covered updates to Inductor, distributed training, dynamic shapes and Apple Silicon support, the event page lists Shulga with Thinking Machines Lab.³⁶²⁴

Work and contributions

• PyTorch (Apple MPS/Metal backend, build and CI) Open-source code

  Listed as a PyTorch core maintainer and as a maintainer of the Apple M1/MPS/Metal backend and Build + CI; the PyTorch CODEOWNERS file names Shulga as an owner of the MPS and Metal source directories and of the CODEOWNERS file itself. The repository is under a BSD-style license.¹⁵⁷

Timeline

1. Apr 28, 2025Presented the PyTorch 2.7 release live Q&A as a core maintainer at Meta.³
2. Jan 28, 2026Presented the PyTorch 2.10 release live Q&A with Andrey Talman and other speakers.⁶
3. Mar 31, 2026Presented the PyTorch 2.11 release live Q&A with Andrey Talman.²
4. Sep 17, 2026Presented the PyTorch 2.14 release live Q&A, listed with Thinking Machines Lab.⁴

Other documented roles

Roles the sources document that are not confirmed as current.

• Software Engineer, responsible for PyTorch releases and continuous integration, Meta³²

Where they work

• Thinking Machines Lab 13 people in the wiki
• PyTorch 11 people in the wiki

Related people

• Sam GrossAlso at Thinking Machines Lab
• Soumith ChintalaAlso at Thinking Machines Lab
• Horace HeAlso at Thinking Machines Lab
• Woosuk KwonAlso at Thinking Machines Lab
• Stephen RollerAlso at Thinking Machines Lab
• Ian BuckChips and hardware · ML frameworks and libraries
• Abubakar AbidDeveloper tools · ML frameworks and libraries
• Piotr BialeckiML frameworks and libraries · Chips and hardware
• Thomas J. FanML frameworks and libraries · Developer tools
• Clément DelangueDeveloper tools · ML frameworks and libraries