I’m a final-year Bachelor’s student at the University of Groningen studying Artificial Intelligence, with a focus on Machine Learning & Data Science. Here I'm also a Teaching Assistant for Introduction to Machine Learning. I wrote my Bachelor’s thesis on the existence of Double Descent in Reinforcement Learning and am eployed at the German startup descript.de, where I'm working on computer vision for textile defect detection in collaboration with TU Dresden.
Other than that, I've built:
- Time-series electricity price prediction for the Dutch state energy provider Gasunie
- Clark Bott: My Openclaw Agent
- NN from Scratch: A CNN implementation from scratch (imports numpy), tested on GTSRB
- PDF Prompt Injector: Adds invisible (size 0) text to PDFs that LLMs can be influenced by, deployed on Streamlit. Useful for resumes 🙃
- Picmash: a Django-based Elo rating system for photo contest ranking
- CoverRoomBets: A python utility to periodically rebalance a market-cap weighted index portfolio made for Cover's trading competition. Won 1st place!
- Mapart Carbon Dating: A utility to estimate when a specific mapart was made on Dominion
- Fibonacci Optimization: Some just-for-fun attempts to calculate the highest Fibonacci number in 0.1s on my machine
- Numista API Interface: Built for my dad to estimate the total value of a numismatic collection
- Dominion Staff Timeline: Staff team size and composition timeline visualization, see Colab deployment
- Whitelist Data Analysis: Analysis of ticket workload distribution among Dominion staff team, writeup
- Lots of course-specific repos that are public where plagiarism rules allow
I also enjoy programming puzzles:
- Advent of Code Solutions 2023--present, including my solution management script. 2nd in 2026 on University leaderboard.
- Cover Crash & Compile solutions for Cover's event, won 3rd place; 2nd in coding.
Email: richitdd @ the standard google mail domain (hint: it's gmail)
Always interested in opportunities and anything cool or neat!
Thanks for stopping by 🤠