Open to opportunities

Kristina Shagina

@kristinashagina

Message

Data Scientist & ML Engineer delivering production ML systems that impact business.

Georgia

Message

What I'm looking for

I seek senior ML/Data roles where I can ship production ML, drive business impact, mentor teams, and build robust MLOps and analytics practices.

I am a Data Scientist and Machine Learning Engineer with over six years of hands-on experience building and deploying end-to-end machine learning systems across recommender engines, computer vision, NLP, and LLM agents.

I prioritize business metrics over model metrics, combining exploratory analysis, feature engineering, and rigorous experimentation to produce solutions that move company KPIs.

My work includes shipping production recommender systems that increased click-through by ~20%, building computer vision solutions with 89–99.9% accuracy, and automating ML pipelines to reduce model update cycles by 40–50%.

I also introduced ML tooling and data culture at organizations, implemented ETL and MLOps practices (MLflow, DVC, Airflow, CI/CD), and trained stakeholders to embed data-driven decision making.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

Neuroom

Mar 2026 - Present (1 month)

Designed a production-grade LLM-based system for automated homework grading, using a RAG architecture with evaluation pipelines and human-in-the-loop feedback, achieving ~85% accuracy, 92% recall, and 91% precision per-question.

Python Pydantic LLM RAG

Data Scientist / Data Analyst

DocsInBox

Dec 2023 - Mar 2026 (2 years 3 months)

Built ML and analytics capabilities from scratch in an organization with no existing data culture or ML expertise
Built churn models and reused the underlying features/models for broader product and customer analytics.
Designed and deployed a fuzzy-matching model for product name normalization across vendors (string similarity + embeddings), reducing manual mapping effort.

Python SQL ETL Churn Modeling Fuzzy Matching Airflow Data Cleaning

Machine Learning Engineer

Data Monsters

Oct 2022 - Jun 2023 (8 months)

Built an NLP system for document security audits using transformer-based models (BERT-like), including information extraction and verification pipelines, reducing manual review effort and improving audit consistency.

Machine Learning Engineer

Yarus

Sep 2021 - Oct 2022 (1 year 1 month)

Developed production-grade recommender systems (collaborative filtering + ranking models) for personalized news feeds, increasing CTR by ~20%.
Trained and evaluated deep learning models for filtering and ranking content.
Introduced ML experiment tracking (MLflow, DVC) to boost team productivity.

Recommender Systems Deep Learning PyTorch MLFlow DVC A B Testing Model Evaluation Ranking PostgreSQL Clickhouse MongoDB Redis scikit learn Faiss Elasticsearch

Data Scientist

Heineken

Jan 2021 - Sep 2021 (8 months)

Developed algorithms to optimize merchandising operations reducing costs by 5–10% and built an MVP to minimize low-quality beer production using predictive features and data collection.

predictive modeling Time Series Analysis Data Collection Optimization Python A B Testing

Data Scientist / ML Engineer

Kinoplan

Sep 2019 - Jan 2021 (1 year 4 months)

Developed a computer vision system detecting people, belongings, and empty seats with 89–99.9% accuracy, integrated models with apps/POS, and automated ML pipelines to reduce model update cycles by 40–50%.

Computer Vision Object Detection Python Docker POS Integration