Software Engineer (AI)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Software Engineer (AI): Building and maintaining distributed training infrastructure for generative video models with an accent on performance optimization, scalability, and developer experience. Focus on profiling end-to-end training pipelines, eliminating bottlenecks in compute and storage, and working close to the metal to maximize accelerator efficiency.
Location: Jerusalem (Onsite)
Company
is an AI-first company creating next-generation content creation technology, including the LTX-2 generative video model and the Facetune creative brand.
What you will do
- Build and maintain distributed training frameworks, including orchestration, checkpointing, and fault tolerance.
- Profile end-to-end training runs to eliminate bottlenecks across compute, memory, and data pipelines.
- Collaborate with researchers to translate model ideas into efficient, production-ready training code.
- Own shared codebases with a focus on correctness, readability, and long-term maintainability.
- Work close to the metal by tuning collective communication and exploiting hardware features.
Requirements
- Strong software engineering fundamentals with proficiency in Python and modern C++.
- Proven experience with performance profiling, optimization, and reasoning about resource contention.
- Deep understanding of distributed systems and debugging issues at scale.
- Familiarity with Kubernetes or similar container orchestration environments.
- Ability to understand systems end-to-end without treating layers as black boxes.
Nice to have
- Working knowledge of accelerator architectures like GPU or TPU.
- Experience with accelerator programming frameworks such as JAX/Pallas, Triton, CUDA, OpenCL, or Metal.
- Prior exposure to ML training pipelines or related pet projects.
Culture & Benefits
- On-site chef-led restaurants and an in-office bakery.
- Daily door-to-door shuttles and Car-to-go subscriptions.
- Access to cutting-edge tools, professional workshops, and learning platforms.
- Collaborative environment focused on craft, challenge, and innovation.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →