Назад
Company hidden
2 дня назад

Lead/Senior Research Scientist – Agentic Reinforcement Learning

Формат работы
hybrid
Тип работы
fulltime
Грейд
senior/lead
Английский
b2
Страна
UK/Switzerland
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Lead/Senior Research Scientist – Agentic Reinforcement Learning (AI/LLM): Developing agentic reinforcement learning methods and reliable training signals for frontier-scale language models performing complex, multi-step tasks in professional domains with an accent on reward design, post-training, tool use, and expert-grounded evaluation. Focus on designing controlled ablations, diagnosing reward and data failures, building scalable training systems, and leading research toward published and deployed results.

Location: Hybrid in Zug, Switzerland, or London, United Kingdom

Company

hirify.global provides specialized software, trusted content, and insights for legal, tax, accounting, compliance, government, and media professionals.

What you will do

  • Lead a research agenda in agentic reinforcement learning and reward design, turning emerging methods into a roadmap for frontier-scale language models.
  • Invent dense, reliable, and cost-efficient training signals for real-world tasks with scarce, delayed, or partly subjective ground truth.
  • Design controlled ablations and diagnostics to distinguish algorithmic effects from data, reward, and system artifacts, and transfer results into production models.
  • Collaborate with product and engineering teams, academic research institutions, subject matter experts, and Research Engineers building large-scale asynchronous training infrastructure.
  • Publish as a lead or senior author at top-tier venues and represent hirify.global Labs through seminars, lectures, conferences, publications, and technical assets.
  • Mentor scientists and engineers and guide research from ideas to trained models, published results, or deployed outcomes.

Requirements

  • PhD in machine learning, computer science, or a related discipline, plus 3+ years of research experience after the PhD in industry or academia.
  • Strong publication record at top-tier venues on RL for LLMs, reward modelling, reasoning, agents and tool use, or preference optimisation.
  • Hands-on experience post-training LLMs with reinforcement learning or preference optimisation using advanced PyTorch.
  • Proven experience mentoring or supervising researchers, engineers, or students and guiding work to published or deployed results.
  • Ability to lead a research agenda from idea to published, reproducible results and design controlled ablations while diagnosing reward, data, and optimisation failures.
  • Excellent written and oral communication, a self-driven attitude, and comfort with ambiguity in fast-paced environments.

Nice to have

  • High-impact publications or recognised community roles, such as area chair, or widely adopted methods or benchmarks.
  • Experience with NeMo-RL, verl, OpenRLHF, TRL, and multi-turn tool environments such as NeMo-Gym-style environments.
  • Experience designing LLM-as-a-judge or rubric-based reward pipelines and measuring their reliability.
  • Experience with Mixture-of-Experts or long-context models in RL settings.
  • Experience with knowledge-intensive or professional-domain tasks in legal, financial, or scientific settings, particularly grounding and citation.

Culture & Benefits

  • Flexible hybrid work environment with digital and physical collaboration.
  • Work from anywhere for up to 8 weeks per year, alongside flexible work-life balance policies.
  • Career development through continuous learning, skills development, and Grow My Way programming.
  • Benefits include flexible vacation, two company-wide Mental Health Days, Headspace access, retirement savings, tuition reimbursement, incentive programs, and wellbeing resources.
  • Two paid volunteer days annually and opportunities for pro-bono consulting and ESG initiatives.
  • Inclusive, globally distributed work environment focused on customer impact, learning, collaboration, accuracy, fairness, and transparency.

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 →