FIRAS MOSTAFA // PORTFOLIO

FULL-STACK
& AI ENGINEER

I build intelligent applications end-to-end — from Flutter mobile clients to Django backends and production ML & LLM systems.

TECHNICAL STACK

01

AI & Machine Learning

LangChain, RAG pipelines, PyTorch, and gradient-boosting ensembles (XGBoost / LightGBM / CatBoost) — trained and shipped in real products.

02

Backend Engineering

Django & DRF with async architecture: WebSockets (Channels), Celery, Redis, PostgreSQL, and Docker.

03

Cross-Platform Mobile

Flutter with BLoC and Clean Architecture — polished, production-grade apps from a single codebase.

SELECTED WORKS

01

Ursa

An AI-driven health companion and smart triage platform. Users describe symptoms in natural language; the system analyzes them against habits and wearable vitals, then triages the case and generates a documented medical report. Built two ML models — a diagnosis ensemble (246K+ records, 721 conditions, 98% Top-5) and a triage model (94% critical-case recall) — plus a medical RAG pipeline with strict safety guardrails.

Flutter Django PyTorch LangChain RAG PostgreSQL Docker
02

TasteCraft AI

A full-stack AI culinary and nutrition platform: smart recipe discovery, automated calorie tracking, personalized meal planning, and a voice-enabled AI chatbot — with full Arabic/English (RTL/LTR) support.

Flutter Django REST JWT LangChain Docker
View on GitHub →
03

AgriSphere

A machine-learning application that serves farmers: guidance on optimal planting times and diagnosis of plant diseases from data, helping improve yield and reduce crop loss.

Flutter Machine Learning Laravel MySQL React
View on GitHub →

LET'S CONNECT

Open to remote Full-Stack / Backend / AI roles.

Email firassmostafa15@gmail.com