Waleed Raza Portfolio

Waleed Raza

development portfolio

Featured Projects

Fully developed and/or works in progress:

MMAmania – iOS App

The Ultimate AI-Powered MMA Platform

Built with: Dart, Flutter, Swift, Firebase, Supabase, Python (spaCy, NLTK, PyTorch, Pyro, TensorFlow), AWS EKS, PostgreSQL, FastAPI, Apache Spark, Ray

Click to preview

View Project GitHub Demo

Pokerswound

Production-Ready Poker with Real Entropy

Built with: TypeScript, Node.js, PyTorch, TensorFlow, React, Firebase, Supabase, PostgreSQL, Redis, Docker, AWS EKS

Click to preview

View Project GitHub

StockSurveyor

Enterprise-Grade AI-Powered Trading Intelligence Platform

Built with: MongoDB, Express, React, Node.js, Python, Golang, TensorFlow, scikit-learn, LangChain, Redis, Supabase, Firebase, Flutter (in progress)

Click to preview

🎯 Revolutionary Trading Intelligence

Institutional-Grade Technical Analysis: Elliott Wave, Fibonacci, Harmonic patterns with 25+ professional indicators
Real-Time Multi-Source Data: 15+ data sources including Alpha Vantage, Yahoo Finance, Reddit, Twitter, Google News
Advanced Sentiment Intelligence: NLP-powered news, social media, and earnings analysis with bias modulation system
Supply/Demand Zone Analytics: Institutional-level zone identification and scoring with options flow intelligence

🏗️ Enterprise Architecture

Multi-Stack Technology: React web app + Flutter mobile, Node.js/Express API gateway, Python ML services, Golang high-performance services
Data Layer: MongoDB for document storage, Redis for caching, Supabase for authentication, Firebase for push notifications
AI/ML Infrastructure: LSTM price prediction (85%+ accuracy), LightGBM risk scoring, CNN-based pattern detection (90%+ accuracy)
Real-Time Processing: WebSocket integration with <50ms latency, Apache Kafka for event streaming, Redis-powered caching
Professional Interface: TradingView integration, custom indicators, multi-timeframe charts, dark/light themes

🔧 Advanced Features

Bias Modulation System: Dynamic GPT-4 prompting with bullish/neutral/bearish tones for user-controlled narrative shaping
Elliott Wave Analysis: AI-powered 5-wave impulse and ABC correction identification with ML-powered forecasting
Fibonacci & Harmonic Analysis: Advanced retracement levels, extension projections, golden ratio analysis
Professional Trading Interface: Interactive candlestick charts, custom indicators, chart annotations, real-time updates

View Project GitHub

AgentYeager

Local-First Framework for Building LLM-Powered Agent Swarms

Built with: Python, SQLite, Ollama, PyYAML, asyncio, FastAPI, Redis, Docker, LangChain, PyObjC (for macOS native UI integration)

Click to preview

🎯 Enterprise-Grade Agent Framework

100% Local-First Design: Complete privacy with no cloud dependencies, all processing on local infrastructure
Multi-Agent Swarms: Coordinate autonomous agents with shared state management and load balancing
Advanced Memory System: SQLite-backed persistent memory with reflection capabilities and task chaining
Modular Skill System: Pluggable skills for any automation task with marketplace distribution
RAG Pipelines: Local embedding lookup and retrieval-augmented generation with hybrid retrieval strategies

🏗️ Architecture & Components

Agent Configuration: YAML-based declarative agent definitions with hierarchical settings and dynamic skill loading
Ollama Integration: LLaMA3/Mistral models for streaming multi-turn reasoning and local LLM inference
Task Coordination: Async task queue with Redis state management, fault tolerance, and auto-scaling
CLI Interface: `agentflow run`, `agentflow create-agent` commands for seamless local-first agent control
Native UI Integration: macOS native interface via PyObjC with real-time dashboards and task monitoring

🔧 Advanced Features

Live Commands: Real-time agent execution like `agentflow run my_agent_config.yaml 'Summarize today's Reddit top NVIDIA news'`
Skill Development Framework: Custom automation capabilities with rate limiting, validation, and versioning
Task-Aware Execution: Real-time monitoring, intelligent scheduling, and performance optimization
Memory Management: Encrypted storage with intelligent context retrieval, compression, and automatic backup
Security & Privacy: Encrypted SQLite databases, secure token management, comprehensive security auditing

View Project GitHub

Hand-Recognition Shortcuts Application

Advanced Hand Gesture Recognition System

Built with: Python, OpenCV, MediaPipe, PyTorch, TensorFlow, Tkinter, scikit-learn

Click to preview

🎯 Revolutionary Hand Gesture Recognition

Real-time Hand Tracking: Advanced MediaPipe integration for precise 21-point hand landmark detection at 30fps
Machine Learning Pipeline: Custom CNN-LSTM hybrid model for improved gesture recognition accuracy with 94% precision
Adaptive Learning System: Self-improving system that learns from user interaction patterns and personalizes recognition
Multi-Modal Input: Hand gestures with optional Whisper speech command fallback for enhanced accessibility
Intelligent Fallback System: Ensemble methods using scikit-learn SGD classifier and XGBoost for robust performance

🏗️ Technical Architecture

Computer Vision Pipeline: OpenCV processing with advanced image preprocessing, landmark analysis, and multi-resolution processing
ML Engine: CNN-LSTM hybrid with transfer learning, attention mechanisms, and adversarial training for robust recognition
User Interface: Tkinter-based real-time feedback with gesture overlay, calibration interface, and performance metrics
System Integration: Flask server with SocketIO, pyautogui for desktop automation, pygetwindow for advanced window management
Research-Grade Implementation: Novel algorithms for temporal gesture recognition with comprehensive analytics

🎮 Gesture Library & Features

Core Gestures: Index finger (copy), three fingers (select all), pinky finger (paste), thumb directions (desktop switching)
Advanced Gestures: Multi-finger combinations, dynamic gestures, custom mappings, context-aware actions
User Calibration: Automatic personalization based on individual gesture patterns with real-time feedback
Performance Monitoring: Live accuracy and latency monitoring with comprehensive analytics dashboard
Educational Impact: Downloaded by 267+ students in the George Mason CS department, used in CV lab demos

View Project GitHub

Petform

All-in-One Pet Management App

Built with: Dart, Flutter, Firebase, Firestore, Supabase (legacy), Auth0, PostgreSQL, and Google Cloud Functions

Click to preview

View Project GitHub

About Me

Software developer

Currently focused on ML modeling, LLms/transformer models, NLP stance and sentiment algorithms, Computer Vision applications and models, Distributed system architecture and design, Full-Stack web and IOs Development, Data pipeline and platform engineering

Role: Software Engineer Intern @ Freddie Mac.
Education: Computer Science BS

Let's Connect

Interested in collaborating on interesting projects or discussing opportunities? I'm always open to:

Collaborations: Open source projects, hackathons, research
Work: Full-time roles, internships, contract deve
Networking:chats, discussions

Email via Gmail LinkedIn GitHub Collaboration Form Resume Portfolio

Skills

Technologies and frameworks I work with

Waleed Raza

Waleed Raza

Featured Projects

MMAmania – iOS App

The Ultimate AI-Powered MMA Platform

🚀 Core Features

🔧 Technical Architecture

📊 Data Coverage & Impact

Pokerswound

Production-Ready Poker with Real Entropy

🎯 Production-Ready Features

🔧 Technical Architecture

🎮 Game Features & Innovation

StockSurveyor

Enterprise-Grade AI-Powered Trading Intelligence Platform

🎯 Revolutionary Trading Intelligence

🏗️ Enterprise Architecture

🔧 Advanced Features

AgentYeager

Local-First Framework for Building LLM-Powered Agent Swarms

🎯 Enterprise-Grade Agent Framework

🏗️ Architecture & Components

🔧 Advanced Features

Hand-Recognition Shortcuts Application

Advanced Hand Gesture Recognition System

🎯 Revolutionary Hand Gesture Recognition

🏗️ Technical Architecture

🎮 Gesture Library & Features

Petform

All-in-One Pet Management App

🤖 AI-Powered Pet Assistant

📊 Comprehensive Health Tracking

🌐 Community & Shopping Features

About Me

Software developer

Let's Connect

Skills