Hi, my name is
Jake Funk
I'm a full-stack developer.
Take a look at some of the things I've worked on, and If you have an idea for a project, I'd love to chat!
About me
I'm currently working as a software engineer at the Johns Hopkins University Applied Physics Lab developing physics simulation and modeling tools. I graduated magna cum laude from UMBC with a degree in Bioinformatics and Computer Science in 2023, and started my Master's degree in Computer Science at Johns Hopkins University in 2024.
As a developer, my main professional goal is to build projects that have a positive impact. I really enjoy being a part of teams that are creating tools to help the world become more efficient, safe, inclusive, and connected.
Projects
Please, feel free to take a look at some of the cool things I've developed.
-
Clueless
Next.js, shadcn/ui, FastAPI, Docker, Github Actions, Google Cloud Platform
Clueless is a simple take on the classic tabletop game Clue, that can be played online. I built the frontend using Next.js 14 as the meta-framework and used shadcn/ui for the components. The server is written in python using the FastAPI framework, which makes use of the openAPI standard and pydantic for data validation. I used uv for the python environment management.
The CI/CD pipeline runs all the tests for the project, then builds a docker image of the server and deploys it automatically to Google Cloud Run.
Click here for the code or Click here for the live demo
-
Dialdetect.com
Nextjs, Vercel, Twilio, Supabase, Stripe
Dial Detect is a side project I built with a friend of mine in the marketing industry. It’s a tool for affiliate marketers and lead gen sales people to automate their direct inbound dialing (DID) services.
The frontend is all Next.js and the backend is the same, with connections to supabase for the DB, twilio for voice over IP (VoIP) and stripe for payments.
-
Breast Cancer Diagnosis Classifier
Python, SciKit-Learn, XGBoost, Seaborn
Using popular machine learning/data science libraries, I trained several classifier models on UCI’s Wisconsin Breast Cancer dataset in order to compare how accurately different types of models performed.
I used Seaborn to create modern histograms of this dataset’s feature distributions, as well as ROC curves for analyzing model performance. After determining that my XGBoost classifier was the most accurate, I tuned the model’s hyperparameters for further optimization, resulting in a model with over 97% accuracy and over 96% precision (cross-validated on 10 folds).
-
PlanGex
React, MUI Data Grid, Node + Express, SQLite
PlanGex is a genetic database CRUD application I built while volunteering as a research assistant and developer for UMBC’s Lobo Lab. PlanGex exposes a REST API and a simple frontend that lets lab members search through, sort, organize, and export data from the lab’s planarian worm genetics database.
PlanGex also helps the lab members find publications that mention specific genes, as well as visualizations of predicted gene expression patterns from known outcomes in published experiments.
Let's Connect
I love collaboration and I am currently looking for new opportunities! Please feel free to reach out to me whichever way is best for you, I am actively checking all of the contact methods listed here, so please feel free to send me an email, message, connection request, DM, etc.