Computer Science undergraduate deeply passionate about building and shipping AI systems — not just training models, but taking them all the way from raw data to deployed, production-ready applications.

I focus on end-to-end AI pipelines: dataset curation, model architecture, training, evaluation, optimization, and deployment behind clean REST APIs. Every project I build is designed to work in the real world — fast, accurate, and maintainable. I'm also OCI Certified across AI, Data Science, and Autonomous Database, which grounds my ML work in production-grade cloud infrastructure.

"I don't just train models. I build systems around them."

• 🔭 Currently building end-to-end computer vision and MLOps pipelines
• 🌱 Currently deepening my knowledge of Agentic AI, LLMOps, and Kubernetes-based model serving
• 🎯 Actively working toward final year project & exhibition milestones
• 💬 Ask me about YOLO, transfer learning, FastAPI model serving, or OCI AI services
• ⚡ Fun fact: I debug faster with coffee than without ☕
• 📫 Reach me at YOUR_EMAIL

• Python (Advanced — ML, DL, scripting, APIs)
• Java (OOP & DSA foundations)
• SQL (Querying, schema design, optimization)

• Exploratory Data Analysis (EDA) & Feature Engineering
• Supervised Learning — Classification, Regression, Ensemble Methods
• Model Evaluation (AUC, F1, Precision, Recall, Confusion Matrix)
• NLP — Text preprocessing, Sentiment Analysis, Summarization
• Information Retrieval Systems (TF-IDF, Inverted Index, Boolean Retrieval)
• Kaggle competitions & real-world datasets

• TensorFlow & Keras — model design, training loops, callbacks, fine-tuning
• PyTorch — custom architectures, research experiments, flexible pipelines
• Transfer Learning & Fine-tuning (EfficientNet, ResNet, MobileNet, VGG)
• YOLO — object detection, custom dataset training, real-time inference
• Image Classification — CNNs, data augmentation, class imbalance handling
• Medical Imaging — preprocessing, clinical metric optimization
• OpenCV — real-time vision pipelines, frame processing, camera integration

• Oracle Cloud Infrastructure (OCI) — AI Services, Data Science, Autonomous Database
• Google Cloud Platform — model training and deployment (free-tier optimized)
• Cost-aware architecture for solo/student-scale AI deployments
• Exploring Kubernetes for scalable model serving

• FastAPI — high-performance REST APIs for ML model serving
• Flask — lightweight API backends and web interfaces
• Docker — containerizing AI applications for reproducible deployment
• GitHub Actions — CI/CD pipelines, automated testing and deployment
• Streamlit — rapid ML demo and data app interfaces
• Clean API architecture — route design, request validation, error handling

• PostgreSQL, MySQL, SQLite, MongoDB, OCI Autonomous Database
• SQL querying, schema design, indexing

• Linux (CLI, environment management, scripting)
• Git & GitHub (version control, clean repo practices)
• Jupyter Notebooks & Google Colab (training & experimentation)
• Kaggle (competition datasets, GPU-accelerated training)

End-to-end AI system helping blind and visually impaired users navigate their environment independently.

• Real-time YOLO-based object detection for obstacle identification and environment understanding
• Custom-trained model optimized for low-latency mobile inference
• Voice-guided feedback system with spoken environment descriptions
• FastAPI backend handling model inference, request routing, and response delivery
• React Native (Expo) mobile client consuming the AI backend
• Focused on real-world usability, accessibility, and inclusive design

End-to-end deep learning pipeline for medical image classification — built with patient safety as the primary constraint.

• Transfer learning with EfficientNet, fine-tuned on breast ultrasound imaging data
• Full training pipeline: preprocessing, augmentation, class imbalance handling, training, evaluation
• Clinically relevant evaluation: AUC, Precision, Recall, F1 — optimized to minimize false negatives
• Deployed as a production-ready web application via FastAPI with image upload interface
• Demonstrates full AI systems lifecycle: data → model → API → application

Real-time computer vision system for ID card localization and institutional verification.

• Custom YOLO model trained on a purpose-built dataset for ID card detection
• Real-time camera feed processing for on-the-spot verification
• FastAPI backend for inference requests with fast response times
• Full end-to-end CV pipeline: data collection → annotation → training → deployment

A comprehensive collection of ML & DL implementations built throughout coursework and independent research.

• Covers: Classification, Regression, CNNs, NLP, Object Detection
• Includes: Malaria Cell Classification, House Price Predictor, and more
• Each project includes full training code, evaluation, and FastAPI deployment
• Well-documented notebooks suitable for learning and reference
• 📂 View Repository

• 📌 AI Systems Developer / ML Engineer
• 📌 Computer Vision & Medical AI
• 📌 Model Training, Fine-tuning & Optimization
• 📌 MLOps & Production Model Deployment (Oracle Cloud + Docker)
• 📌 End-to-End AI Pipeline Engineering