Research Engineer (NLP)

About The Company

MIL Team (Machine Intelligence Laboratory, MIPT) is a Skolkovo resident and the machine intelligence lab of the Moscow Institute of Physics and Technology. The team has delivered 150+ AI projects, employs 60+ researchers, and has spun off multiple ventures, including receiving the Huawei Best Cooperator Award. Focus areas include applied AI: generative computer vision, model compression, NLP, deep learning in signal processing, and classic machine learning.

Responsibilities

• Propose and test hypotheses using modern approaches in NLP and LLM tasks.
• Read research papers and reproduce results.
• Work with retrieval, RAG, AI agents, reasoning, and long-context pipelines.
• Adapt open-source models to applied scenarios.
• Conduct experiments and compare approaches based on quality, latency, and inference cost.
• Quickly build MVPs and research prototypes.
• Use AI tools to accelerate research and engineering work.
• Prepare technical materials summarizing project outcomes.

Requirements

• Proficiency in Python, Git, and Linux.
• Experience with the ML/NLP stack: PyTorch, transformers, pandas, numpy.
• Understanding of modern LLM and Transformer architectures.
• Hands-on experience with NLP tasks: classification, retrieval, summarization, QA, agents, and related areas.
• Ability to design and conduct reproducible experiments.
• Capability to progress from problem definition to implementation and solution justification.
• Openness to AI-native workflows: Cursor, Copilot, Claude Code, agent pipelines, and AI-assisted research and development.

Nice-to-have

• Experience with RAG, vector databases, reranking, and evaluation.
• Experience with inference optimization, quantization, vLLM, TensorRT-LLM, and similar tooling.
• Open-source contributions.
• Experience writing scientific papers.
• Interest in leading projects or launching new research directions.

Working Conditions

• Fully remote work from anywhere in the world.
• Collaboration with a high-caliber team offering significant learning opportunities.
• Highly dynamic environment with rapid decision-making and hypothesis testing.
• Opportunities for vertical growth and movement across ML domains.

Format: Participation through the MIL Talent Pool. You will be matched to a project when a suitable task arises.