Open to opportunities
Syed Muhammad Abdullah Masood
Next.js · Django · FastAPI · PyTorch — from model to production
Selected work — keep scrolling
Production platforms with Next.js and React on the frontend and Django/DRF or Express APIs over PostgreSQL and MongoDB behind them.
Applied ML and computer vision — realtime detection with YOLOv8, MediaPipe, and OpenCV, LLM/vision pipelines, and end-to-end data science.

Local-first résumé shortlisting in Rust — parses a folder of PDFs, embeds them with quantized ONNX models, and ranks candidates against a job spec with a 3-stage pipeline: Qdrant vector search, cross-encoder reranking, and BM25 lexical scoring.

Predicts Falcon 9 first-stage landing success from launch parameters using scraped flight data and classical ML.
The same product shipped everywhere — native desktop builds with Tauri and mobile apps with Flutter.

The TaskForge core packaged as a native Tauri desktop client with system-tray integration.

Dark-themed Flutter music player for Android — local library plus Last.fm-powered discovery, MP3 downloads, background playback with full media-control integration, a native equalizer, and rich playback settings.
Low-level C and C++ — a from-scratch document database, memory allocators, arbitrary-precision arithmetic — plus interactive 3D games in Unity.
I'm a full-stack engineer who builds AI-powered products end to end — realtime computer vision, LLM and vision pipelines, and the apps that ship them. The through-line is one product, everywhere it needs to live: Next.js and Django on the web, FastAPI for AI services, Tauri on the desktop, and Flutter on mobile.
Full toolkit
Languages
Web & Backend
ML/AI & Computer Vision
Data Science
DevOps & Tools
Final-year student. Coursework: AI, Web Development, Database Systems, Operating Systems, Applications of Virtual Reality.
12-course program across the full data science lifecycle — Python, SQL, analysis, visualization, ML — with an applied capstone.
Loading repos…
Have a project in mind or just want to say hi? Drop me a message and I'll get back to you.
ElectroStore client is a Next.js/React e-commerce frontend for browsing and managing electronics. Features Firebase authentication, product browsing by category, admin dashboard, and protected routes. Sessions persist across page reloads. Integrates with a REST API backend. Built with Ant Design for a response.
Noir Player is a lightweight Flutter music player supporting background playback, local media querying, and a responsive UI. It integrates audio_service, just_audio, and on_audio_query for a seamless experience. Features include a tabbed library, a "Now Playing" screen, and background audio.
A Unity-based 3D recreation of Angry Birds featuring physics-based slingshot mechanics, multiple levels, bird abilities, and WebGL compatibility for seamless browser gameplay
Analyze SpaceX Falcon 9 rocket launches to predict first-stage landing success using Python, Pandas, machine learning, and interactive dashboards. This project covers data collection, wrangling, geospatial analysis, predictive modeling, and visualization to uncover insights into launch outcomes and cost efficiency.
Forecasts smartphone battery degradation using sensor data from a Samsung device. After cleaning and analyzing time-series features like temperature, CPU usage, and voltage, an LSTM model predicts battery percentage and generates short-term forecasts, showing strong accuracy but limited real-world variability.
Analyze Pakistan’s crop yields under changing climate using Python, Pandas, and machine learning. The project covers data wrangling, visualization, and modeling to uncover how temperature, rainfall, and evapotranspiration affect agricultural productivity