Projects
Research projects and production systems spanning quantitative finance, computer vision, NLP, and full-stack development.
Bayesian Optimization for Basket Trading
Developed quantitative asset allocation algorithm using Bayesian Optimization to maximize out-of-sample returns while mitigating overfitting inherent in traditional statistical backtesting. Implemented acquisition functions (Expected Improvement, Upper Confidence Bound) for efficient hyperparameter search across portfolio configurations. Python, NumPy, SciPy, Bayesian Methods.
Provenance — UNESCO Hackathon Winner 🏆
Created award-winning deepfake detection system for digital media verification. Implemented multimodal analysis combining audio spectral features and visual artifact detection to identify AI-generated misinformation. Media literacy platform helping people see truth, spot fakes, and build trust online. NLP, Deep Learning, Computer Vision, JavaScript.
WasteVision — Real-Time Waste Classifier
Engineered production-grade object detection model achieving 0.801 mAP for waste classification. Optimized inference pipeline for 10ms latency using TensorRT and model quantization. Deployed via Streamlit interface for real-time smart recycling applications. PyTorch, YOLOv5, Streamlit, Computer Vision.
Hybrid Customer Support Chatbot
Architected two-tier conversational AI system combining rule-based Rasa intents with GPT-3.5 fallback generation. Achieved 90% response accuracy and reduced average resolution time by 40% through contextual dialog management. Rasa, GPT-3.5, NLP, FastAPI.
Market Intelligence Platform
AI-powered solution for entrepreneurs and startup founders to analyze market data, identify business opportunities, track industry trends, evaluate competitors, forecast demand, and assess risks by combining real-time market data with advanced AI analysis. TypeScript, AI/ML, Real-time Analytics.
Telecom Churn Prediction
Telecom churn prediction and analysis project using machine learning. Built predictive models to identify customers likely to churn, enabling proactive retention strategies. Python, Scikit-learn, Pandas, Jupyter.
Real-Time Face Emotion Detection
Real-time face detection Streamlit-based web application for server deployment. Detects and classifies facial emotions using computer vision and deep learning models. Python, OpenCV, Streamlit, Deep Learning.
Owlet — AI Study Companion
An AI study companion and mentor designed to guide students using the Socratic method to foster academic integrity and independent learning. TypeScript, AI/ML, Education Tech.
Disaster Relief AI Platform
AI-powered web platform that analyzes real-time data (satellite imagery, weather data, social media posts) to identify disaster-hit areas and predict the extent of damage for emergency response. AI/ML, Satellite Imagery, Real-time Data.
CyberSafe-PK — Security Platform
Crowdsourced security platform connecting ethical hackers with organizations. Features vulnerability reporting, automated triage, and secure communication protocols. Incubated at National Incubation Center. JavaScript, Django, PostgreSQL, Security.
Symphonic Stories
An interactive system that translates voice into musical and visual art, capturing emotions through stories and mapping them into visual and auditory forms of storytelling. Python, Audio Processing, Visualization.
Jenkins CI/CD Pipeline
Complete CI/CD setup using Jenkins demonstrating automated testing, building, and deployment. The application provides basic mathematical operations through a REST API. Python, Jenkins, REST API, DevOps.
Digital Image Processing & OCR
Essential files for Digital Image Processing (DIP), along with an OCR Project focused on Visual Pattern Recognition. Includes test images, key DIP concepts, and MATLAB code files. MATLAB, Image Processing, OCR.