💡 Merging robust full-stack architecture (MERN & Python/Flask) with cutting-edge Deep Learning & Computer Vision to craft intelligent, real-time, production-grade applications.

B.Tech Computer Science & Engineering graduate from Trident Academy of Technology (Graduated May 2026) and alumnus of Sainik School Bhubaneswar.

With 2 months of hands-on AI/ML internship experience across CTTC MSME and 1stop.ai, alongside end-to-end full-stack software development skills, I bridge the gap between intelligent deep learning models and scalable web platforms.

• 🚀 Full-Stack Proficiency: React.js, Node.js, Express.js, Flask, MongoDB, MySQL, WebSockets, Socket.io, Docker.
• 🧠 AI/ML & Computer Vision: TensorFlow, Keras, OpenCV, DeepFace, Dlib, Scikit-learn, NLP, CNN architectures.
• 📦 Proven Track Record: Architected and shipped 3 production-grade full-stack applications & deployed operational AI systems.

Jan 2025 – Jun 2025

• Real-Time Attendance System: Engineered an operational face recognition attendance system using CNNs, TensorFlow, Keras, and OpenCV, directly integrated with a MySQL database for live automated logging.
• Custom Object Detection: Built object detection pipelines using transfer learning and custom data augmentation techniques (rotations, brightness shifts, flipping) to maximize model generalization.
• Production Integration: Collaborated in a 4-member cross-functional team to embed deep learning models into desktop client software for live operational workflows.

Jul 2025 – Dec 2025

• NLP Text Classification: Designed a multi-class NLP pipeline featuring tokenization, TF-IDF vectorization, word embeddings, and deep neural network classifiers.
• Landmark & Image Recognition: Developed end-to-end CNN image classification models in TensorFlow/Keras from raw data ingestion to evaluation (Precision, Recall, F1-Score).
• ML Pipeline Optimization: Implemented regularization strategies (Dropout, Batch Normalization, Early Stopping) to minimize train-validation variance.

Tech Stack: Python · Flask · MySQL · DeepFace · OpenCV · Dlib · Bootstrap · Stripe · Docker

• Biometric Authentication: Prevents impersonation at login using DeepFace real-time facial verification against registered student profiles.
• Live AI Proctoring Engine: Built with OpenCV & Dlib's 68-point facial landmark model to continuously track gaze direction, multi-face presence, and absences with timestamped anomaly logs.
• Multi-Format Exam & Built-in Compiler: Supports MCQs (with CSV upload & negative marking), subjective tests, and an in-browser coding compiler supporting 15+ languages (C, C++, Java, Python, Node.js).
• Monetization & Deployment: Integrated Stripe API for professor credit top-ups; fully Dockerized with MySQL database and Flask-Session management.

Tech Stack: React.js · Node.js · Express.js · MongoDB · Socket.io · Redux Toolkit · JWT

• Instant Board Synchronization: Trello-style drag-and-drop Kanban boards powered by @hello-pangea/dnd and Socket.io WebSocket broadcasting for instant multi-user board updates without refreshing.
• Optimistic UI Updates: Instant client-side state mutation with seamless UI rollback on API network failure.
• Workspace Security & Audit Trail: Granular RBAC permissions with JWT middleware and background activity trail logging stored in MongoDB.

Tech Stack: React.js · Redux Toolkit · Node.js · Express.js · MongoDB · Chart.js · JWT

• Triple Portal Architecture: Dedicated, security-scoped portals for Admin, Teacher, and Student roles with Express middleware protection.
• Rich Assessment Creator: Dynamic question banks, automated evaluation logic, and custom exam timers built with React rich-text editing.
• Analytics & Anti-Cheat: Real-time detection of browser blur events, tab switches, and window focus changes paired with Chart.js analytics dashboards for class metrics.

Tech Stack: Python · OpenCV · CNN · TensorFlow · Keras · MySQL

• End-to-End Pipeline: Live video feed ingestion $\rightarrow$ Haar Cascade face detection $\rightarrow$ CNN feature verification $\rightarrow$ Automated MySQL database attendance logging with timestamps.
• Production-Adopted: Developed during internship at CTTC MSME and deployed for actual operational use.

Tech Stack: Python · TensorFlow · Keras · Scikit-Learn · Kaggle

• Full NLP pipeline handling text cleaning, tokenization, stop-word removal, TF-IDF vectorization, word embeddings, and deep neural network training evaluated with F1-Score & AUC-ROC metrics.

• 🏅 AI/ML Training Program: CTTC MSME, Bhubaneswar (2025)
• 🏅 AI/ML Internship Certification: 1stop.ai (2025)
• 🏅 Full-Stack Web Development (MERN Stack): Self-Directed / GitHub Portfolio (2024 – 2025)

aditya_profile = {
    "current_status": "B.Tech CSE Graduate (Completed May 2026)",
    "seeking_roles": [
        "Full-Stack Developer", 
        "AI/ML Engineer", 
        "Front-End Developer", 
        "Software Engineer"
    ],
    "currently_building": [
        "Intelligent MERN Applications", 
        "Edge ML Computer Vision Models"
    ],
    "key_strengths": [
        "End-to-end Web Architecture (MERN & Flask)",
        "Biometric & Computer Vision Systems (OpenCV, DeepFace, TensorFlow)",
        "Real-Time WebSockets & Optimistic UI Updates",
        "Containerization (Docker) & Clean Production Code"
    ]
}