I'm a Master's student in Data Science at Aalto University with a background in Computer Science. My work focuses on solving challenging mathematical and statistical problems through data analysis, optimization, and machine learning.

• M.Sc. in Data Science - Aalto University (In Progress)
• B.Sc. in Computer Science - Aalto University
• Strong foundation in mathematics and statistics
• Bachelor's thesis on graph-enhanced recommender systems

I specialize in applying rigorous mathematical and statistical methods to analyze and optimize systems. My interests span:

• Machine Learning - Developing predictive models and exploring their implications
• Optimization - Solving computational challenges with mathematical approaches
• Knowledge Graphs - Leveraging graph-based methods for recommendation systems

I'm pragmatic about technology choices—I evaluate and learn whatever tools best suit the problem at hand.

Languages & Runtimes

Data Science & ML

Web & Infrastructure

Tools

A production-ready Telegram bot for tracking physical activities and running guild-based fitness competitions. Features a comprehensive point system based on the 2024 Physical Activity Compendium, with interactive wizards for activity logging and a React-based analytics dashboard.

Key Features:

• Multi-step conversation flows for intuitive activity logging
• Guild-based competition system with leaderboards and rankings
• Real-time statistics dashboard with data visualizations
• Production deployment on Kubernetes with Talos OS
• Infrastructure as Code with Terraform and Flux CD

Technologies: TypeScript, Bun, Telegram Bot API, PostgreSQL, React, Vite, Tailwind CSS, Kubernetes, Docker, Terraform, Hetzner Cloud

Investigated fatality rate patterns across aircraft manufacturers using hierarchical Bayesian models in R with brms. Our analysis revealed persistent manufacturer-level differences in safety outcomes despite standardization efforts, with partial pooling models achieving the best predictive performance.

Technologies: R, brms, tidyverse, bayesplot, ggplot2

Built regression models to assess whether Apple smartphones are overpriced relative to their hardware specifications. Tested multiple algorithms (linear regression, decision trees, random forest, neural networks) and found that newer iPhone models carry larger premiums, with all Apple devices priced above hardware-justified levels.

Technologies: Python, scikit-learn, pandas, seaborn

Bachelor's thesis comparing traditional recommendation algorithms with knowledge graph-based approaches. Explored knowledge graph embeddings and graph neural networks, analyzing the trade-offs and optimal use cases for each methodology.

Technologies: Python, Graph Neural Networks, Knowledge Graph Embeddings

Full Stack Developer Intern - Droppe (via Aalto University)
Developed an analytics platform where I designed the data schema, optimized data pipelines, and built UI components including a no-code filter system for chart creation. Led documentation efforts and pitched the project to finals of the course quality awards (Top 3).

Technologies: TypeScript, React, Node.js, Google Cloud Platform, BigQuery

I'm seeking internship opportunities where I can apply my analytical and technical skills to solve meaningful problems. I'm particularly interested in roles that involve data science, machine learning, or mathematical optimization.

Feel free to connect with me on LinkedIn. I'm always open to discussing interesting problems, collaboration opportunities, or potential roles.

When I'm not analyzing data, you'll find me training for triathlons or on the basketball court.