Staff Research Engineer (AI)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Staff Research Engineer (AI/LLM): Developing and deploying industry-leading conversational AI models for enterprise support with an accent on model training, orchestration, and evaluation. Focus on improving agent reliability, instruction following, and solving complex long-horizon task completion challenges in production.
Location: On-site in New York City
Salary: $325K – $425K + Equity
Company
is a conversational AI platform empowering brands to deliver concierge customer experiences through advanced AI agents.
What you will do
- Lead research and engineering efforts to enhance production capabilities, including instruction following, retrieval, and memory.
- Build and iterate on end-to-end models and pipelines optimized for quality, efficiency, and user experience.
- Partner with platform and product engineers to integrate new models into production systems.
- Translate ambiguous research ideas into clear, iterative milestones and strategic roadmaps.
- Mentor other researchers and engineers while establishing technical best practices for applied research.
Requirements
- 8+ years of experience in AI/ML engineering or research.
- Proven experience post-training and deploying LLMs in production environments.
- Fluency in Python and modern ML tooling for training, evaluation, and data pipelines.
- Track record of taking research prototypes to reliable, measurable production impact.
- Must be based in New York City for on-site work.
Culture & Benefits
- Flexible "take what you need" vacation policy.
- Comprehensive Medical, Dental, and Vision insurance for employees and their families.
- Retirement planning including 401K options.
- Parental leave and fertility/family building benefits through Carrot.
- Daily lunches and snacks provided in the office.
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 →