обновлено 2 дня назад
Member of Technical Staff, Machine Learning (AI)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
Текст:
TL;DR
Member of Technical Staff, Machine Learning (Python/PyTorch/JAX): Building and improving production ML components across data, training, evaluation, and inference with an accent on model reliability, real-world task completion, and GPU-based systems. Focus on fine-tuning models, designing robust data pipelines and evaluation systems, and debugging latency, cost, reliability, safety, and production issues.
Location: Hybrid in London, United Kingdom
Company
is building proactive AI-native applications that help users manage conversations, errands, organisation, and workflows with minimal prompting.
What you will do
- Build and improve ML components across data, training, evaluation, and inference.
- Fine-tune and adapt models within larger production systems running on GPUs.
- Implement evaluation and testing to understand model behaviour and improve reliability.
- Build and maintain pipelines for real-world and synthetic data.
- Debug model issues, performance problems, and production incidents.
- Collaborate with senior ML engineers and product teams to ship iterative improvements.
Requirements
- Strong foundations in machine learning and modern neural architectures.
- Hands-on experience training, fine-tuning, or deploying ML models.
- Production-quality Python development.
- Experience or familiarity with PyTorch or JAX and production ML systems running on GPUs.
- Ability to work through ambiguity, learn quickly, and grow ownership over time.
- Ability to work under production constraints including latency, cost, reliability, and safety.
Culture & Benefits
- Work in a small, high-talent-density, hands-on team.
- Make decisions collectively while moving at a rapid pace.
- Balance high-quality delivery with continuous learning and iteration.
- Work directly with real production systems and user feedback from day one.
Hiring process
- Applications are evaluated by technical team members.
- Expect three, and no more than four, interviews conducted virtually and/or onsite.
- Decisions are made promptly after the interview process.
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 →