Hire the Best Data Scientists
in Estonia

I build Python-first Data Science and AI systems that turn messy data into reliable predictions. I can deliver notebooks, dashboards (streamlit/Gradio/ Tabeau/ Power BI), and production-ready APIs. What you can hire me for • Classical ML and forecasting: EDA → feature engineering → training/tuning → evaluation, plus time series forecasting with interpretable trend/seasonality. • NLP with Transformers: BERT/RoBERTa fine-tuning, dataset prep, evaluation, error analysis, and practical deployment. • LLM Multi Agent systems: RAG pipelines, tool calling, structured outputs, agent orchestration, and output validation so it’s reliable. • Production delivery: FastAPI endpoints, SQL-backed pipelines, Docker, CI/CD-oriented workflows, reproducibility and clean code. Honors & Awards • Finalist, MumbaiHacks2025 (India) — the largest Agentic AI hackathon (20,000+ registrations) [2025] • Completed Data Analyst training program, Kreativstorm (Python, R, IBM SPSS) [2024] • Awarded Performance Stipend, University of Tartu, for excellent academic results [2021] • INSPIRE Fellowship (India) — National Merit Award for top 1% students in Natural Sciences [2016–2021] Tech stack (Python-centered) - Python: Pandas, NumPy, SciPy, scikit-learn, PyTorch, TensorFlow, FastAPI - LLM tooling: LangChain, LangGraph, HuggingFace, Google ADK, Google AI Studio, OpenAI, Gemini - Data: PostgreSQL, SQLite, complex joins, CTEs, window functions - MLOps and infra: Docker, GitHub Actions, DVC, cloud deployments (AWS/GCP), reproducible pipelines - Visualization: Matplotlib, Streamlit, Tableau, Power BI How I work - I start by clarifying the objective and success metrics - I deliver in milestones with quick prototypes first, then harden into production quality - I prioritize reproducibility, documentation, and evaluation so the system stays maintainable Background - PhD candidate (Particle Physics), trained in rigorous quantitative modeling and building high-precision algorithms - Strong communication: I’m comfortable translating technical results into clear decisions for stakeholders If you share your dataset, current pipeline, and the outcome you want (prediction, LLM assistant, dashboard, or API), I can suggest a clean plan and deliverables for the first milestone.

Over 7 years of experience in the IT industry, specializing in Cloud-native Data systems and AI architecture, with a proven track record in designing scalable and secure infrastructure on various cloud platforms to solve complex challenges, including real-time fraud detection, agentic automation, and risk identification. Expert in deploying Large Language Models (LLMs) and machine learning ensembles to enhance shareholder decisions and regulatory compliance.

I'm a Microsoft Certified Data Scientist with 7 years of experience in Data Science and Machine Learning, and 20+ years in software development. I have worked with industry leaders such as the European Central Bank, Santander Bank, Deutsche Bahn, and RTL Deutschland, delivering high-impact solutions. How I Can Help You: - Knowledge Extraction – Extract and structure data from your internal knowledge base. - Intelligent Search & Hybrid RAG Applications – Enhance data retrieval with semantic, text, and graph search technologies.   Technical Expertise:   Programming: Python, Java, Javascript. Data Extraction: OCR and VLM pipelines, pipeline evaluation and LLM fine-tuning. Gen AI Development: LangChain, Hugging Face. Databases: SQL (PostgreSQL), NoSQL (Elasticsearch, Opensearch, MongoDB, Neo4j, ArangoDB). LLM Traceability: LangFuse, Phoenix Cloud: AWS, Azure, GCP."

My name is Thomas, Highly skilled Full-Stack Developer and Data Scientist with over 6 years of experience in building robust web applications and implementing advanced data analysis techniques. Proficient in modern web technologies and frameworks, and experienced in machine learning, deep learning, computer vision, and natural language processing. Passionate about leveraging technology to solve complex problems and deliver high-quality, scalable solutions.