LLM Research Engineer

We are seeking an experienced Data Scientist with a passion for large language models (LLMs) and cutting-edge AI research. In this role, you will design and prototype data preparation pipelines, collaborating closely with data engineers to transform your prototypes into scalable production pipelines, design and implement a state-of-the-art evaluation and benchmarking framework to measure and guide model quality, and do end-to-end LLMs training. You will work alongside top AI researchers and engineers, ensuring our models are not only powerful but also aligned with user needs, cultural context, and ethical standards.

What you will do

• Curation of datasets for pre-training, supervised fine-tuning, and alignment;
• Research and develop best practices and novel techniques in LLM training and evaluation pipelines;
• Collaborate closely with data engineers, annotators, linguists, and domain experts to scale data processes, define evaluation tasks and collect high-quality feedback.

Qualifications and experience needed
Education & Experience:

• 3+ years of experience in Data Science or Machine Learning, preferably with a focus on NLP;
• An advanced degree (Master’s or PhD) in Computer Science, Computational Linguistics, Machine Learning, or a related field is highly preferred.

GenAI & NLP Expertise:

• Practical experience with fine-tuning LLMs / VLMs models;
• Hands-on experience with modern NLP approaches, including embedding models, semantic search, text classification, sequence tagging (NER), transformers/LLMs, RAGs.

ML & Programming Skills:

• Strong experience with deep learning frameworks such as PyTorch or JAX for building models;
• Ability to write efficient, clean code and debug complex model issues.

A plus would be
Advanced NLP/ML Techniques:

• Applied experience using Reinforcement Learning in NLP / LLM settings;
• Prior work on LLM safety, fairness, and bias mitigation;
• Experience generating and curating synthetic datasets for Supervised Fine-Tuning (SFT), including quality control and scaling considerations.

Research & Community:

• Publications in NLP/ML conferences or contributions to open-source NLP projects;
• Active participation in the AI community or demonstrated continuous learning (e.g., Kaggle competitions, research collaborations) indicates a passion for staying at the forefront of the field.

MLOps & Infrastructure:

• Hands-on experience with containerization (Docker) and orchestration (Kubernetes) for ML, as well as ML workflow tools (MLflow, Airflow);
• Experience in working alongside MLOps engineers to streamline the deployment and monitoring of NLP models;
• Experience with cloud platforms (AWS, GCP, or Azure) and big data technologies (Spark, Hadoop, Ray, Dask) for scaling data processing or model training is a plus.

Problem-Solving:

• Innovative mindset with the ability to approach open-ended AI problems creatively;
• Comfort in a fast-paced R&D environment where you can adapt to new challenges, propose solutions, and drive them to implementation

What we offer

• Office or remote — it’s up to you. You can work from anywhere, and we will arrange your workplace;
• Remote onboarding;
• Performance bonuses;
• We train employees with the opportunity to learn through the company’s library, internal resources, and programs from partners;
• Health and life insurance;
• Wellbeing program and corporate psychologist;
• Reimbursement of expenses for Kyivstar mobile communication.