Hi, I'm
M.Sc. Data Science · Software & Data Engineer
Master of Data Science with a background in software engineering and data engineering. I conduct full data science projects end-to-end — from data pipelines and classical ML to LLM-powered systems and production APIs.
Master of Data Science based in Finland, specializing in end-to-end AI and analytics solutions. I bridge the gap between raw data and production-ready applications by combining robust software engineering with advanced machine learning. My focus is on delivering reliable and maintainable systems. I leverage modern AI coding assistants to accelerate development while maintaining strict quality control and rigorous code review standards for production safety.
My interest in data science sparked during my computer science bachelor's degree, motivating me to focus heavily on math and statistics to build a strong foundation for data science. Getting my master's degree was a rewarding journey that taught me how to apply deeply theoretical concepts to real, hands on tasks.
Outside of work and personal projects, creativity is a huge part of my life. Whether I am producing music, playing the piano, DJing or playing video games and tabletop RPGs, I find that my creative mindset directly helps me tackle challenges in unique ways.
📍 Finland
Supervised and unsupervised methods — regression, classification, random forest, SVM, clustering, dimensionality reduction, and model selection with rigorous evaluation.
RAG pipelines, multi-step agents, tool-use architectures with LangChain, and comfortable use of agentic coding tools with rigorous code review practices.
Production FastAPI services, full stack development, REST APIs, async job queues, and analytical Jupyter notebooks for business insight derivation.
ETL pipelines, data collection and cleansing, stream processing with Kafka, vector databases, PostgreSQL, and cloud platforms — Fabric, Azure, Snowflake, Databricks.
Exploratory data analysis, statistical storytelling, Jupyter notebooks, data visualisation, and translating findings into clear business recommendations.
Master's thesis conducted for Telia Finland. Built ML models to predict maximum B2B telecommunications revenue potential for Finnish companies, enabling data-driven target marketing. Managed end-to-end in Microsoft Fabric: data collection, cleaning, feature engineering, and model training.
Trained a Double Deep Q-Network (DDQN) reinforcement learning agent to play Super Mario Land. Covers environment wrappers, replay memory, target network updates, and reward shaping to achieve stable game-playing behaviour.
Collects historical and real-time stock news via the Alpaca API and runs FinBERT sentiment analysis on each article. Users select a time frame and get a breakdown of positive, neutral, and negative news for any ticker. News stored in Azure Blob Storage and cleaned in Databricks using a medallion architecture.
Production-ready FastAPI template used as the foundation for most of my Python software projects. Includes RAG with pgvector, persistent chat, JWT auth, async background jobs, and pluggable LLM support.
Platform for uploading and hosting ML models and datasets, with agentic sentiment analysis on uploaded data. Supports model management, dataset exploration, and AI-driven insight generation.
Portfolio site built with FastAPI and Jinja2 — serves both the HTML frontend and a public REST API. Deployed on Render via Docker with GitHub Actions CI/CD.