Lead R&D Engineer (Agentic AI)
Мэтч & Сопровод
Для мэтча с этой вакансией нужен Plus
Описание вакансии
TL;DR
Lead R&D Engineer (Agentic AI): Lead the research and development of advanced NLP, LLM, and Agentic AI solutions moving from experimentation to production impact with an accent on multi-agent architectures, tool-use integrations, and retrieval-augmented generation pipelines. Focus on architecting scalable training, fine-tuning, deployment workflows, and establishing MLOps best practices for responsible AI systems.
Location: Poland, Ukraine (Remote/Office)
Company
Global IT consulting company with R&D-focused AI initiatives shaping technical direction and innovation strategy.
What you will do
- Define and lead development of advanced NLP, LLM, and Agentic AI solutions ensuring high quality and real-world applicability.
- Own end-to-end R&D lifecycle from research to production-ready architectures delivering business impact.
- Collaborate with product, engineering, and research teams to align AI with priorities.
- Design multi-agent systems, tool integrations, and RAG pipelines.
- Architect scalable workflows using cloud and distributed environments.
- Mentor AI engineers, establish evaluation and monitoring practices, and communicate to stakeholders.
Requirements
- 7+ years in AI/ML, including 3+ years leading R&D and production projects
- PhD/Master’s in ML/CS or equivalent senior experience
- Deep expertise in NLP/LLMs: training, fine-tuning, RAG, evaluation
- Experience with Agentic AI, multi-agent orchestration, tool integration
- Advanced Python, C++/Java; PyTorch, Hugging Face, LangChain, LlamaIndex, vector DBs
- Cloud (AWS/GCP/Azure), distributed training, MLOps, Docker, Kubernetes
- Strong leadership, mentoring, communication, ethical AI commitment
Culture & Benefits
- Equal Opportunity Employer committed to diversity and inclusion.
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 →