Skip to content
View therealemmanueloppong's full-sized avatar
🎯
Focusing
🎯
Focusing

Highlights

  • Pro

Block or report therealemmanueloppong

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Typing SVG

πŸ‘¨πŸΎβ€πŸ’» Emmanuel Oppong

Computer Science Student β€’ Builder

"Ship it, then make it better."


⚑ About Me

I'm a junior Computer Science student at Grambling State University (3.95 GPA) who builds practical software for real people. I'm looking for a Summer 2027 software engineering internship.


πŸ”¨ Projects

  • πŸ“± ZipMath (try the live demo) is a game-style mental-math app with lessons, XP, daily streaks and 26 levels across Basic, Intermediate and Advanced tiers. Built with React Native (Expo) and TypeScript. Every practice question is generated in code, so the answers are always correct, and a web demo deploys to GitHub Pages through GitHub Actions. Next: automated tests, accounts and sync with Supabase, then a store release.
  • 🀝 InterviewPair pairs students for mock technical interviews by topic, level and time. A Spring Boot API with JWT login, hashed passwords and a tested matching rule, plus a React and TypeScript frontend, served from one Docker image with 30+ automated tests and CI on GitHub Actions. It isn't deployed yet. Java and Spring Boot are new to me, so this is where I'm learning them.
  • ⚑ GPU Robot Lab is GPU kernels written from scratch in Python (Triton) and benchmarked against PyTorch on an NVIDIA T4, with an interactive explainer. I tuned a matrix-multiply kernel 3.4x (1,096 to 3,727 GFLOPS), found that its compiled code uses no tensor cores (the likely cause of the remaining gap), and built a MuJoCo MJX robot simulation up to 126x faster than one CPU thread, with 90+ automated tests in CI.
  • πŸ“Š Stock Price Prediction (AI4ALL Ignite) is a 3-person team project that predicts next-day stock decreases from financial news headlines, using FinBERT sentiment features and 20M+ stock price records. The Random Forest reached 54.4% accuracy against 50.8% for the Logistic Regression baseline, and I wrote the evaluation script.

🌱 Open Source


🧭 Up Next

  • πŸ” Availability and live matching for InterviewPair, then deploying it.
  • 🧠 A spaced-repetition engine for ZipMath.
  • 🌱 Code contributions, not just docs, to open source projects.

🀝 Community

  • Academic Chair, ColorStack Grambling: I lead weekly data structures and algorithms workshops for 90+ members.
  • AWS Student Builder Group Leader (starting Nov 2026)
  • πŸ… Awards: 2x Dean's List, 4x Academic Excellence Scholarship

βš”οΈ Tools I Use

Learning and using in current projects


πŸ“« Let's Connect


Pinned Loading

  1. jiaying-l/AI4All_Project jiaying-l/AI4All_Project Public

    Jupyter Notebook 1

  2. zipmath zipmath Public

    TypeScript