Here I list noteworthy projects that I have built (or added to) either alone or with a team.

Spotlight Projects

🗄️ RAG/Vector DB/BERT Fine Tune

In this project, I worked on implementing a RAG system. This included fine tuning BERT on a kubernetes cluster, comparing vector databases such as Annoy and FAISS, and using LangChain to build the pipeline.
I worked with one other student and I worked on English data while they worked on German data.
Disclaimer: The GitHub is built like research code, other projects do a better job of indsutrial structure.
Presentation
Vector DB Notebook
RAG Notebook
GitHub

Working at the BHT’s Calgo Lab with Sebastian Jaeger I added to the tab-err, realistic errors in tabular data, package on PyPi.
Wrote basic tests, developed a high level api, and modified some other error types as well as worked on bug fixes in python.
PyPi
GitHub

Implementation of the K-Means algorithm to cluster high population cities based on geographic coordinates.
I wrote this in R Markdown for fun.
Report

In this project, I worked with a group of 4 other students to create a report on the CIA World Fact Book.
We were guided through the analysis by 5 research questions, the last two were done by me.
Report
GitHub
CIA World Fact Book

In this project, I worked with a team of 2 other students to implement various time series forecasting models.
We utilized kubernetes, docker, and VSCode Remote SSH to prototype models on the BHT Berlin’s V100 & A100 GPUs.
I scraped the website of my friend’s weather station to acquire the data and implemented two models.
Presentation
GitHub

Details the importance of data visualization by an exploratory analysis of Anscombe’s Quartet in R.
Written in R Markdown
Report

I read the book Machine Learning Kurz und Gut in German and pointed out a few errors to the author.
Shows some proficiency in German written communication.
Github

In this project, I setup messaging with the Telegram API, GitHub actions, and yfinance to ping my phone each day before the US Stock Exchange opens with information about the S&P 500 and crypto.
A future direction I am interested in is having LLMs summarize news articles and send a daily report to my phone from a curated list of sources.
GitHub

This is a presentation for a team project to analyze data from a 2006 AOL USA search query dataset which we hosted as an Exasol database on a Virtual Machine.
We were required to augment the database with another dataset, so we chose weather events in the USA during 2006.
The skills developed were: SQL, Python, OLAP, ETL, Data Visualization, Tableau, and EDA.
Presentation

In this project, I loaded data from various German government sources about the population and charging stations in various postal codes in Berlin.
A heatmap overlay of a map of Berlin, subdivided by postal code is presented with options for viewing population, number of charging stations, and estimated new station demand.
Finally, a suggestion box functionality allows users to input suggestions attached to a given postal code.
Streamlit Page

In this application I loaded 2020 IRS data and used linear interpolation of tax brackets to compute the estimated median income of various zip codes in the United States.
Streamlit Page
Data

In this project, I implemented (alone, in NumPy, without the use of LLMs) a customizable feed forward neural network that can have an arbitrary number of layers and hidden units.
This was done from the absolute basics, just matrix & vector operations to implement forwardpropagation, backpropagation, and minibatched stochastic gradient descent.
GitHub

In this project, I worked with a team of 3 other students to apply and compare several different statistical learning methodolgies in R on a real estate dataset from Kaggle.
Report
GitHub

This project was the culminating work of my senior research project in my bachelor’s of computer science.
I worked with a team of 3 people to build an applied deep learning model in Python using PyTorch, NumPy, Pandas, and Astropy to predict parameters of binary star systems.
Report
GitHub

In this project, I worked with one other student to analyze a set of time series concerning energy economics data.
During this project, we worked with faculty in the department of business and economics as well as the department of computer science.
The main techniques employed were clustering of time series and sequence mining. Traditional techniques for forecasting were not explored.
Report
GitHub

In this project, as we worked through the textbook: Operating Systems Concepts 10th Edition, we had to individually implement (in KPL & Assembly) a toy operating system, the Blitz OS.
Due to academic integrity constraints I cannot publish my solution for this but I found the course instrumental in developing a deepened understanding of how operating systems work.
Book
Blitz OS Description
Blitz OS GitHub

This was a midterm live coding question which I did before the advent of ChatGPT.
We were given only 50 minutes to complete the program but I managed to get full points including extra credit.
GitHub

In this class project, I implemented a client server architecture that allowed for the communication of messages using TCP/IP and UNIX sockets in C.
Assignment Description
GitHub

In this class project, I implemented a UDP load balancing server using UNIX sockets in C.
The assignment description is on the GitHub but in MS Word…
GitHub

In this assignment, I followed a lab to implement a text editor in Java. While not the most involved project, it is still a fun one.
GitHub