Welcome to my GitHub! I'm dedicated to pushing the boundaries of Machine Learning, Deep Learning, and Audio Processing through innovative projects and real-world applications.
π FineSecure - Financial Fraud Detection & Cybersecurity System
Advanced ML-powered cybersecurity solution for fraud detection, analyzing 6.3 million financial transactions using Logistic Regression and Random Forest classifiers. Identifies fraudulent patterns and security threats in mobile money platforms across CASH-IN, CASH-OUT, DEBIT, PAYMENT, and TRANSFER transaction types with high precision to combat payment fraud and protect financial ecosystems.
π€ EchoRobber
Developing a real-time emotional state detection system from voice recordings using cutting-edge deep learning architectures.
π§ Advanced Emotion Recognition Models
Building high-accuracy audio processing systems with CNNs, spectrogram analysis, and hybrid neural architectures.
π AI-Powered Web Applications
Creating interactive ML applications using Streamlit, featuring chat analyzers and intelligent calculators with predictive capabilities.
π³οΈ Pothole Detection System - Carnival Vision Challenge
YOLOv11-based real-time pothole detection system developed for the Carnival Vision Challenge competition on Kaggle, organized by Computer Coding Club MNNIT. Features interactive Streamlit web application with configurable detection parameters, test-time augmentation, and comprehensive visualization of detected road damage.
π€ MCP Servers Collection with Agent
Comprehensive Model Context Protocol (MCP) implementation featuring 9 specialized servers orchestrated by a LangGraph ReAct agent. Integrates Weather API, Gmail, Google Workspace (Drive & Calendar), GitHub, SQLite, OpenStreetMap, mathematical computation, web search, and robotics club management through unified LLM-powered interface using Groq's Llama 3.1 8B Instant model.
- Machine Learning & Deep Learning - Specializing in audio processing, NLP, and real-time inference systems
- Open Source Contributions - Active contributor to TensorFlow, PyTorch, and Streamlit ecosystems
- Real-Time Audio Analysis - Expert in emotion detection, noise reduction, and speech enhancement systems
- Computer Vision & Object Detection - YOLO architectures, real-time detection systems
- Financial Technology & Fraud Detection - ML-powered security solutions for fintech
- Agent-Based Systems - LangChain, LangGraph, and MCP implementations
- ποΈ Advanced Architectures - Transformers, BERT, GANs, and LSTM variations
- π΅ Audio Processing Mastery - Librosa, SpeechRecognition, MFCC, and tonality extraction
- π MLOps & Deployment - Production-ready ML model deployment strategies
- ποΈ Computer Vision - Integrating CV techniques with ML for comprehensive solutions
- π Agent Orchestration - Model Context Protocol and multi-agent systems
- π Python Programming - Core programming, OOP, and advanced data structures
- π§ Deep Learning Frameworks - TensorFlow/Keras and PyTorch implementation
- π€ Audio Processing & Emotion Recognition - Signal analysis and feature extraction
- π Data Analysis & Visualization - Pandas, NumPy, Matplotlib, and feature engineering
- β‘ ML Application Development - Streamlit apps and real-time prediction systems
- π Object Detection - YOLO architectures and real-time computer vision
- π Fraud Detection Systems - ML-powered financial security solutions
- π€ AI Agents - LangChain, LangGraph, and MCP server implementations
ML Algorithms Expertise:
- Supervised Learning: Linear Regression, Logistic Regression, Decision Tree, Support Vector Machine, Regression Tree
- Unsupervised Learning: K-Means Clustering, Hierarchical Clustering, DBSCAN, Gaussian Mixture Model
- Deep Learning: CNN, RNN, LSTM, BiLSTM, GRU, Transformers, GANs
- Computer Vision: YOLO (v8, v11), Object Detection, Image Classification
| Machine Learning & Deep Learning | Audio Processing | Real-time Systems | Computer Vision |
|---|---|---|---|
| CNN, RNN, GRU, LSTM | Speech Emotion Recognition | Live Audio Analysis | YOLO Object Detection |
| BiLSTM, ANN, Transformers | Signal Processing | Real-time Inference | Image Classification |
| BERT, GANs | Feature Extraction | Production Deployment | Video Analysis |
| Fraud Detection | MFCC, Spectrogram | Agent Orchestration | Pothole Detection |
- Scaling Deep Learning Models for production environments
- Advanced Audio Signal Processing techniques and optimization
- Open Source ML/AI Projects with meaningful impact
- Real-time Processing Systems for audio and multimedia applications
- Computer Vision Applications with YOLO and real-time detection
- Financial Security Systems and fraud detection algorithms
- Agent-Based AI Systems using LangChain and MCP