aipythondeep learningreinforcement learningdata pipelinesevaluationagent systemssynthetic data
Вакансия из Hirify Global, списка международных tech-компаний Для мэтча и отклика нужен Plus
Пожаловаться
Зарплата и рынок
развернуть
По международному рынку данных пока нет
На СНГ рынке: $4.7к/мес ($3.8к - $6к)
87
Крутая вакансия
развернуть
Чётко описанная исследовательская роль связывает передовые AI-задачи с production-системами и даёт высокий уровень ответственности, но выбор только из двух офисных локаций ограничивает доступность.
Только офис
Передовые технологии
Полный цикл работы
Сильная команда
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Member of Technical Staff, Applied Research (AI): Building production-scale reinforcement learning, agent evaluation, and synthetic data systems with an accent on alignment, safety, verifiers, and reward signal design. Focus on reproducing frontier research, diagnosing model failure modes, and developing reliable pipelines for long-horizon agentic tasks.
Location: Onsite in Bengaluru, India; Bangkok, Thailand is also listed as a location.
Company
hirify.global is an applied research lab scaling reinforcement learning for safe and reliable agentic AI capabilities.
What you will do
Read current research across reinforcement learning, post-training, agents, evaluations, and synthetic data, selecting ideas to reproduce and develop.
Design long-horizon agentic tasks, verifiers, graders, and execution environments that resist exploitation and provide meaningful partial credit.
Analyze model trajectories and failure modes, translating findings into improved data and reward signals.
Partner with advanced AI research teams on datasets and benchmarks for alignment, safety, and model behavior.
Own long-running research projects and select problems that matter for scaling data toward AGI.
Build and maintain data pipelines, internal tooling, and deep learning implementations with engineering teams.
Requirements
Strong software engineering skills, extensive Python experience, and the ability to own and debug systems end to end.
Experience building evaluations, reinforcement learning environments, agent harnesses, verifiers, or execution sandboxes.
Experience shipping research into production, particularly post-training, distillation, or high-stakes evaluation.
Experience with large-scale agent systems, including orchestration frameworks, tool APIs, distributed execution, observability, and logging.
Ability to establish ground truth in unfamiliar domains by building systems and analyzing their outputs.
Ability to identify misspecified tasks, ambiguous prompts, and graders vulnerable to gaming.
A thoughtful perspective on the societal and safety impacts of deploying general-purpose AI systems.
Nice to have
PhD in computer science, machine learning, or natural language processing, with publications at NeurIPS, ICML, or ICLR; neither is required.
Experience with public benchmarks or task suites such as Terminal-Bench, SWE-bench-style harnesses, or HCAST.
RLHF or RLAIF experience and a perspective on how training choices affect model behavior.
Experience running work through large expert or annotation operations.
Culture & Benefits
Work onsite in Bengaluru, India, with Bangkok also listed as a work location.
Collaborate with researchers, roboticists, founders, operators, and advanced AI research teams.
Full-time employment in the Scaling department.
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 →