Research Engineer (Agentic Models)

TL;DR

Research Engineer (Agentic Models): Building and optimizing multi-step coding agents for IDEs with an accent on SFT, RL-style post-training, and evaluation pipelines. Focus on designing scalable training recipes, developing simulation environments for agent behavior, and shipping high-performance models into production.

Location: Must be based in the Netherlands, Serbia, Germany, Cyprus, UK, Czech Republic, Poland, or Armenia

Компания

hirify.global is a global software development company known for creating professional developer tools and IDEs.

Что делать

• Design and maintain SFT and RL post-training pipelines for multi-step coding agents.
• Train and adapt LLMs for agent workflows including planning and tool use.
• Build simulation environments to measure and compare agent performance on developer tasks.
• Design evaluation frameworks and metrics to analyze agent behavior and close the loop into training.
• Work with large-scale infrastructure for distributed training and data processing.
• Collaborate with research and product teams to ship features into IDEs.

Требования

• Must be based in one of the listed office locations (Netherlands, Serbia, Germany, Cyprus, UK, Czech Republic, Poland, Armenia)
• Hands-on experience training LLMs in research or production settings.
• Proficiency with PyTorch and specialized LLM training stacks like Megatron or NeMo.
• Solid understanding of LLM training fundamentals including distributed training and data pipelines.
• At least 3 years of Python experience in modern ML codebases.
• Product-aware mindset with the ability to own projects end-to-end.

Хорошо, если есть

• Experience with ML orchestrators like Kubeflow, Dagster, or Airflow.
• Familiarity with large-scale data clusters and multi-node GPU training.
• Experience designing evaluation pipelines for LLMs or agents.
• Knowledge of AI agent frameworks and multi-step coding workflows.
• Experience with inference optimization and production model serving.

Культура и преимущества

• Work on cutting-edge AI-powered developer tools used by millions.
• Access to large-scale distributed GPU and MapReduce infrastructure.
• Collaborative environment working with research, product, and infrastructure experts.
• Focus on high-quality, maintainable code and end-to-end project ownership.