Skip to content
View Vali-Hameed's full-sized avatar
4 followers · 1 following

Achievements

Achievement: YOLOAchievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: Quickdraw

Achievements

Achievement: YOLOAchievement: Pull Sharkx2Achievement: Pair ExtraordinaireAchievement: Quickdraw

Highlights

  • Pro

Block or report Vali-Hameed

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

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

Report abuse
Vali-Hameed/README.md 

██╗   ██╗ █████╗ ██╗     ██╗
██║   ██║██╔══██╗██║     ██║
██║   ██║███████║██║     ██║
╚██╗ ██╔╝██╔══██║██║     ██║
 ╚████╔╝ ██║  ██║███████╗██║
  ╚═══╝  ╚═╝  ╚═╝╚══════╝╚═╝

Typing SVG

Profile Views

Website LinkedIn Email

> whoami

CS undergrad at Lancaster University chasing a First — but more importantly, someone who actually ships things.

I build from first principles: no framework cargo-culting, no tutorial-driven code. Across ML systems, full-stack web apps, mobile platforms, and everything in between — I care about taking ideas all the way to deployed, working products.

Currently: building FightPicks 🥊 — a full-stack MMA prediction platform I architected and engineered from the ground up.

Outside the terminal: I'm an MMA fan (Muay Thai & Kickboxing), which probably explains why I gravitate toward problems that require sustained effort and discipline.

> ls ./stack

Languages

Java Python TypeScript JavaScript Dart C C++ Haskell SQL

Frontend / Mobile

Next.js React React Native Flutter Tailwind CSS

Backend

Spring Boot FastAPI Django PostgreSQL MongoDB

Infra / DevOps

Docker GitHub Actions AWS ECS Oracle VPS

> cat ./projects

🚀 Orbital Risk — 🥈 Runner-up, Leeds Hack 2026

A 3D launch-window optimisation system built under hackathon pressure. Real-time satellite orbit and debris field visualisation powered by WebGL and Three.js. A custom risk engine calculates collision probabilities against live debris density data, while an ML-driven weather service generates automated Go/No-Go launch decisions.

Next.js TypeScript React Three.js Python FastAPI Scikit-Learn Pandas

🔗 Live Demo · 📂 Code

🥊 UFC Fight Predictor — Full Microservices Ecosystem

A passion project that became a proper engineering challenge. Three decoupled services working in concert to bring ML-powered fight predictions and community analytics to MMA fans.

How it fits together:

Next.js Frontend  ──REST/JWT──►  Spring Boot API  ──JPA/Hibernate──►  PostgreSQL
                                        │
                                ┌───────┴────────┐
                                ▼                ▼
                         FastAPI ML Service   Python Scraper
                         (Gradient Boosting    (Playwright + BS4
                          on 7,000+ fights,    Oracle VPS · cron
                          66.09% sym acc.)     via APScheduler)

What's inside each service:

🌐 Web App — Spring Boot + Next.js monorepo · 📂 Code

Expand details
  • JWT auth with HttpOnly cookies, refresh token rotation, BCrypt hashing, and JSoup XSS sanitisation
  • Fight Simulator — pit any two fighters from across divisions against each other
  • Community predictions, discussion forums, and fight card polling
  • Real-time leaderboards tracking accuracy, win streaks, and total points
  • Advanced user management: profile editing, 90-day username cooldowns, account deletion with data anonymisation
  • Token-bucket rate limiting via Bucket4j + Caffeine to prevent abuse
  • Admin dashboard for managing users, roles, fights, and ML pre-warming tasks
  • Automated emails via Resend API in production; MailHog locally
  • Schema migrations managed with Flyway
  • Full CI/CD via GitHub Actions: prod deploy on push to main, staging environments on PRs, manual rollback via workflow_dispatch
  • Fully containerised dev stack via Docker Compose — no local Java or Node needed
  • Backend tested with JUnit 5 + Mockito (H2 in-memory DB); frontend with Jest + React Testing Library

Java Spring Boot Spring Security TypeScript Next.js Framer Motion PostgreSQL Flyway Resend Docker Oracle GitHub Actions

🤖 ML Service — Containerised prediction API · 📂 Code

Expand details
  • Gradient Boosting Classifier trained on 7,000+ historical bouts — 66.09% true symmetric accuracy
  • Symmetrised inference: every fight is evaluated as A vs B and B vs A, then averaged to eliminate red-corner bias
  • Per-corner feature engineering (striking rates, takedown defence, etc.) to capture non-linear relationships
  • Hyperparameter-tuned (max_depth=2, n_estimators=300, learning_rate=0.05) to handle skewed target data
  • Served as a containerised FastAPI microservice, deployed on AWS ECS
  • update_dataset.py for local manual dataset backfills; live production scraping handled by the dedicated scraper service

Python FastAPI Scikit-Learn Pandas Docker AWS

🕷️ Scraper Service — Automated live data ingestion · 📂 Code

Expand details
  • Autonomous Python worker scraping live UFC fight cards, events, results, and fighter stats from ufcstats.com
  • Cron-scheduled nightly via APScheduler; /trigger/historical endpoint for one-off backfills
  • Authenticated pushes to the Spring Boot API via X-Scraper-Key header
  • Cached fighter roster (fighters.json) with active/inactive classification by weight class
  • Deployed on Oracle VPS inside a Docker Compose environment

Python FastAPI Playwright BeautifulSoup Docker Oracle

> git log --stat

GitHub Streak Vali's Activity

contribution snake

> cat ./currently

{
  "building"  : "FightPicks — full-stack AI prediction & community platform",
  "studying"  : "Computer Science BSc @ Lancaster University (Year 3)",
  "exploring" : ["distributed systems", "ML serving", "system design"],
  "open_to"   : ["grad jobs", "internships", "interesting problems"]
}

If you're building something ambitious — let's talk.

Pinned Loading

  1. UFC-Fight-Predictor-Website UFC-Fight-Predictor-Website Public

    A full-stack, microservices-based web application for UFC fight analysis. Features community predictions, a fight simulator, real-time leaderboards, and Gradient Boosting ML predictions. Built with…

    Java

  2. UFC-Fight-Predictor UFC-Fight-Predictor Public

    A containerized Machine Learning microservice that predicts UFC fight outcomes using a Gradient Boosting Classifier with 66.09% symmetric accuracy. Built with FastAPI, Scikit-Learn, and Docker, dep…

    Python 1

  3. UFC-Scraper UFC-Scraper Public

    An automated, containerized Python microservice that scrapes live UFC fight cards, event data, and fighter statistics using Playwright and BeautifulSoup. Built with FastAPI and deployed on Oracle VPS.

    Python

  4. akaltemamey/leedshack2026-prototype akaltemamey/leedshack2026-prototype Public

    A 3D launch-window optimisation system combining debris exposure modelling and machine-learned weather risk.

    TypeScript 1

  5. Penalty-Shootout-Predictor Penalty-Shootout-Predictor Public

    A full-stack predictive analytics engine utilizing Bayesian statistics and Monte Carlo simulations to forecast football penalty shootouts. Built with Next.js, FastAPI, and Docker.

    TypeScript

  6. Remote-Terminal Remote-Terminal Public

    A native Windows terminal session manager. Detach from long-running PowerShell scripts and securely monitor or interact with them from your phone over Tailscale.

    JavaScript