Aspiring Full-Stack AI Engineer crafting intelligent, scalable applications at the intersection of Generative AI, Computer Vision, and Full-Stack Engineering.
01 / About
I'm a final-year B.Tech student in Artificial Intelligence & Machine Learning at St. Thomas' College of Engineering & Technology, Kolkata.
My passion lies in the Generative AI ecosystem — from LLM orchestration and RAG architectures to real-time computer vision systems. I'm actively building enterprise-grade AI solutions and deepening my expertise in LangChain, prompt engineering, and AI-native application design.
B.Tech AI & ML
St. Thomas' College of Engineering & Technology
Class of 2026
02 / Skills
03 / Projects
Driver Monitoring System
A real-time, event-triggered driver safety framework engineered to detect drowsiness and complex distraction behaviours with minimal inference latency. Uses a multi-stream architecture that fuses lightweight ocular/vehicular heuristics with heavy deep learning models for edge-hardware deployment.
Implemented a Temporal Convolutional Network (TCN) with Self-Attention to classify time-series distraction patterns.
Multi-stream architecture optimises compute on edge hardware via heuristic + DL model fusion.
Full-stack reporting pipeline with real-time alerts + PDF generation for post-drive analytics.
class AttenDrive:
def __init__(self):
# Multi-stream pipeline
self.ocular = EyeTracker()
self.pose = MediaPipe()
self.tcn = TCN_Attention()
def infer(self, frame):
# Fuse heuristics + DL
eye = self.ocular.process(frame)
body = self.pose.process(frame)
return self.tcn.classify(
eye, body, t=self.ts)
AI Productivity Toolkit
A full-stack AI productivity toolkit with modular endpoints for intelligent document processing, text extraction, and workflow automation. Integrates Tesseract OCR with LLM APIs for context-aware document summarization.
Automated OCR→LLM pipeline transforming unstructured text into structured data.
Precise prompt templates for context-aware extraction and summarization.
Iteratively tested Streamlit UI for seamless GenAI model interaction.
LAN Chat Application
A desktop-based LAN chat application featuring encrypted communication and a multithreaded client-server architecture. Supports multiple concurrent users with real-time messaging, emoji support, and message history encryption.
Multithreaded Java Sockets supporting multiple simultaneous connections.
Swing GUI with emoji support, timestamps, and auto-scrolling.
Message history encryption strengthening data privacy and UX.
04 / Education
Lancaster University, UK
Accepted to pursue advanced studies focusing on cutting-edge AI architectures, machine learning frameworks, and research-driven software development.
St. Thomas' College of Engineering & Technology
Core curriculum spanning Data Structures & Algorithms, Operating Systems, DBMS, Computer Networks, AI & ML, NLP, and Computer Vision.
National Gems Higher Secondary School
Focused heavily on core STEM subjects including Physics, Chemistry, Mathematics, and Computer Science.
National Gems Higher Secondary School
Foundational studies emphasizing analytical thinking, logic, and early exposure to computer applications.
Open Source Contributor
Selected from a highly competitive, global pool of applicants to collaborate with top-tier developers and organizations on critical, open-source AI projects. This selection highlights deep technical proficiency, innovative problem-solving, and a proven ability to contribute at an enterprise-grade level to the real-world software ecosystem.
Smart India Hackathon (SIH)
2023, 2024, 2025
Spearheaded complex, event-triggered edge vision systems, securing finalist and winning placements for immediate real-world impact.
OOFY HACKS 1.0
Overall Winner
Won top honors with an end-to-end full-stack solution integrating advanced Computer Vision for real-time analytics.
AI & ML with Data Science
Workshop · IIT Guwahati
Completed rigorous deep-tech training, deepening understanding of neural architectures and large-scale data pipelines.
IndiaAI Impact Gen-AI
Hackathon Participant
Participated in a nation-wide Gen-AI challenge, exploring RAG architectures for large-scale enterprise solutions.
05 / Contact
I'm actively seeking opportunities in AI engineering, full-stack development, and research roles. Whether you have a project, a position, or just want to connect — my inbox is open.