I’m seeking Senior/Lead ML or LLM/AI Agent roles focused on ML forecasting / LLM integrations, post-training, RAG services and product-impactful AI that accelerates business growth through reliable ML/LLMOps and measurable results.
Roman Prilepskiy
@rprilepskiy
Senior/lead machine learning engineer delivering production AI / ML services, LLM agents, RAG solutions, and post-training with measurable impact.
What I'm looking for
I’m a Senior/Lead Machine Learning Engineer with 10+ years delivering production ML and LLM systems across telecom, classifieds/marketplace, AI product, and public sector environments, including team leadership up to 8 specialists. I specialize in production LLM agents, RAG and search relevance, and SFT/LoRA inference workflows—backed by strong evaluation and LLMOps quality gates.
I drive end-to-end impact from PoC to production-readiness: I took “AI Sales Manager” from demo to a Google ADK MVP for 100 legal-entity clients, improved commercial RAG assistants by up to +58 p.p. MAP@5, and reduced district revenue forecasting RMSE by 45%. I’m focused on building reliable pipelines with guardrails, structured outputs, observability, and synthetic-dialogue evaluation—so product logic uses explicit business statuses and teams can ship faster with confidence.
Experience
Work history, roles, and key accomplishments
LLM Engineer
Avito
Aug 2025 - Apr 2026 (8 months)
Spearheaded “AI Sales Manager” from PoC/demo to a Google ADK MVP for 100 legal-entity clients, achieving 100% KR1 internal acceptance and offline-testing with 9 potential clients vs 1 planned. Implemented production-readiness for tool/assistant flows (structured outputs, guardrails, observability) and built synthetic-dialogue evaluation on Qwen3-235B, including LoRA-based intent coverage for 5 pro
Lead Machine Learning Engineer
MTS AI
Jan 2020 - Jul 2025 (5 years 6 months)
Led RAG and ML product delivery across enterprise assistants, knowledge search, and recommender/forecasting use cases, improving retrieval quality by up to +58 pp MAP@5 (accepted benchmark quality up to 93%) via preprocessing redesign, hierarchical chunking, and Optuna-driven retriever tuning. Built “Happiness Index” and “Smart Rollout” systems, delivering 1.5–2x higher CLV growth, +9 pp churn pre
Middle Data Scientist
TerraLink
Jan 2019 - Dec 2019 (11 months)
Delivered a telecom IT-service request classifier across 65+ categories, achieving F1=0.818 and turning NLP/text classification into a presales-ready ML solution. Created client-facing packaging and value-proposition materials for upsell and lead generation, and presented the work in a public TerraLink webinar.
Data Scientist
BestFitMe
Aug 2017 - Dec 2018 (1 year 4 months)
Improved a car recommendation engine by 5x and ran 10 web A/B tests with 500 users, driving 3x customer-satisfaction growth and supporting like-prediction modeling (F-score=0.75). Built an NLP article-classification pipeline from 15,000 car articles, reaching 0.8 accuracy and 0.7 F-beta.
Data Analyst / Team Lead
Data Analysis Center under Moscow IT Department
Jun 2015 - Aug 2017 (2 years 2 months)
Led the eHealth platform benchmark project as team lead, coordinating 8 analysts and analyzing 39 web services/apps to identify 56 key features for Moscow healthcare strategic planning. Delivered ETL/PL/SQL validation for 20M healthcare records and implemented EMIAS/DWH analytics with fuzzy-logic duplicate detection that resolved 280,000 duplicate records.
Education
Degrees, certifications, and relevant coursework
Skolkovo Institute of Science and Technology
Master of Science, Information Science and Technology
2013 - 2015
Earned an MSc in Information Science and Technology at Skolkovo Institute of Science and Technology (Skoltech) from 2013 to 2015.
Moscow Institute of Physics and Technology
Master of Science, Intellectual Data Analysis
2013 - 2015
Earned an MSc in Intellectual Data Analysis through the joint Skoltech–MIPT program from 2013 to 2015.
Massachusetts Institute of Technology
Summer Program, Innovation Workshop
2013 - 2013
Grade: -
Activities and societies: -
Completed the MIT Innovation Workshop summer program in 2013 as part of the Skoltech–MIT master’s degree partnership.
Bauman Moscow State Technical University
Engineer, Electronic Devices
2007 - 2013
Earned an engineering degree in Electronic Devices at Bauman Moscow State Technical University from 2007 to 2013.
Tech stack
Software and tools used professionally
GitHub
GitLab
Bitbucket
Docker Compose
GitLab CI
NumPy
Pandas
PySpark
MySQL
PostgreSQL
MongoDB
Gmail
Yarn
Telegram
Redis
Jira
Python
JSON
TensorFlow
PyTorch
MLflow
scikit-learn
Keras
Streamlit
NLTK
Gradio
FastAPI
asyncio
SQLAlchemy
Gemini
pytest
Docker
uWSGI
Uvicorn
Airflow
SQL
XGBoost
SciPy
LightGBM
CatBoost
LangChain
Polars
Pydantic
LiteLLM
vLLM
Cursor
ClearML
Black
Ruff
Faiss
Optuna
Rasa
Make
DeepSeek
Windsurf
OpenWebUI
Grok
Qwen
Sentence Transformers
ChatGPT
Availability
Location
Authorized to work in
Salary expectations
Job categories
Skills
Interested in hiring Roman?
You can contact Roman and 90k+ other talented remote workers on Himalayas.
Message RomanFind your dream job
Sign up now and join over 250,000+ remote workers who receive personalized job alerts, curated job matches, and more for free!
