Чётко очерченная senior-роль в крупной продуктовой компании даёт заметную ответственность и работу с реальными AI-системами, но зарплата не указана, а обязательный офис в Таллине ограничивает гибкость.
?Зарплата не раскрыта
Передовые AI-технологии
Ответственность за результат
AI в большом масштабе
?Работа из офиса
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Подключить Plus
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Applied AI Engineer (Operations Automation): Building operations automation systems for customer support, document verification, and fraud detection across a global mobility platform with an accent on LLM evaluation, process redesign, and decision quality. Focus on designing agent and human checkpoints, calibrating LLM judges, running A/B tests, and improving live systems through policies, prompts, tools, and models.
Location: On-site at Bolt's Tallinn HQ, Estonia
Company
Bolt operates a global mobility platform serving more than 200 million customers across 50+ countries and 850+ cities, supported by 4.5+ million partners and 4,000+ employees.
What you will do
Redesign operational processes by deciding which steps should become software, AI agents, or human checkpoints.
Build evaluation suites from real cases and human decisions, turning failures into regression tests.
Develop robust checks for LLM and agent systems, calibrate LLM judges against human labels, and control costs and payouts.
Run improvement loops across policies, prompts, tools, models, and post-training methods.
Own rollout decisions, A/B tests, and the feedback loop between live results and evaluation sets.
Codify successful solutions so other teams can operate them independently.
Requirements
Background in software engineering, machine learning, or data science, with experience shipping and operating systems that call LLMs.
Experience building evaluation suites for LLM or agent systems and using them to make release decisions.
Ability to analyze agent traces, identify decision failures, and improve system behavior.
Experience designing LLM judges, measuring agreement with human labels, and addressing disagreements.
Ability to design and interpret A/B tests, compare offline predictions with live results, and assess cost trade-offs between rules, classical ML, humans, and LLMs.
Clear written English is required; the role is performed on-site in Tallinn, Estonia.
Nice to have
Hands-on post-training experience with RLVR, GRPO, DPO, reward modeling, or LoRA.
Experience with GEPA, DSPy, TextGrad, or similar program optimization frameworks.
RL environment or gym design with programmatic grading.
Operations automation experience in a marketplace, fintech, or trust and safety team.
Strong SQL skills and comfort working in unfamiliar codebases.
Culture & Benefits
On-site collaboration with operations and engineering teams at the Tallinn headquarters.
Professional growth opportunities and career development.
Salary, stock options, and physical and mental wellness benefits.
A one-month paid sabbatical after five years at Bolt.
Annual company events and smaller team gatherings.
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 →