Machine Learning engineer with a bias towards LLM deployment in IK Aktualnaya Mehanika, OOO

Actual Mechanics — company , which is engaged in the development of software in the field of design and calculations of turbomachines, as well as engineering projects aimed at the design of new and modernization of existing turbomachines - steam and gas turbines, industrial compressors, pumps.Requirements:

Education: Graduates and students of departments with specialized education in the field of machine learning.

Professional skills:

• Understanding basic NLP tasks (LLM, embedding models, etc.)
< li>Ability to deploy and maintain ML models
• Knowledge of popular ML frameworks and libraries (e.g., TensorFlow, PyTorch)
• Confidence with Python for scripting, automation, and integration
• Basic understanding of MLOps practices and tools
• Understanding of RAG systems principles
• Git skills

Would be a plus:

• Proficiency in Bash for Linux administration on virtual machines
• Experience with cloud platforms such as AWS, RunPod
• Experience with SQL for data management
• Understanding containerization technologies (Docker)

Responsibilities:

• Training/fine-tuning of Large Language Models (LLMs)
• Optimizing ML models to improve their performance
• Supporting ML models in production, deploying and monitoring
• Ensuring security, reliability and scalability of machine learning infrastructure
• Troubleshooting performance and deployment issues of ML models
• Finding and implementing process best practices MLOps
• Writing documentation and tests for code
• Interaction with ML engineers and Software engineers to optimize the development process

Working conditions:

• official employment;
• high salary level (discussed with each candidate individually based on the results of the interview);
• (Mon-Fri) from 10: 00 to 19:00 or from 11:00 to 20:00;
• fixed weekend Saturday, Sunday;
• paid vacation and sick leave;
• remote work.

If you are interested in our vacancy, do not delay - leave your resume. For questions about the vacancy, call +380996058020 or write to Telegram or Viber (PM -FRI from 10:00 to 19:00).