Engineering intelligent solutions across AI, Machine Learning, Computer Vision, Cybersecurity, and Software Engineering.
I'm a 3rd-year M.Tech (Integrated) CSE student at SSN College of Engineering. I work across the full stack — from systems-level networking tools and IoT hardware to fine-tuned language models and deep learning pipelines.
My research has been recognised at MediaEval 2025 (3rd place, GI Imaging VQA). I'm currently interning as an AI engineer while preparing further publications.
I'm drawn to hard problems — routing anomalies, pixel-level change detection, language in low-resource settings. I care about getting the details right.
Scale 8.0 – 9.0 · S5 is personal best
† School-level assessment only — AY 2020–21 Class X students were declared all-pass; no government board exam was conducted.
Contributing across two parallel tracks at an early-stage data intelligence startup.
Applied BLIP-2 to the Kvasir-VQA-x1 dataset for visual question answering on endoscopy images, generating answers paired with natural-language explanations, attention maps, and confidence scores. Ranked 3rd internationally out of all competing teams.
Awarded the Reliance Foundation Undergraduate Scholarship — a highly competitive, merit-based national scholarship recognising academic excellence and future potential across India.
Ranked 3rd internationally in Subtask 2 of MediaEval 2025 using BLIP-2 on the Kvasir-VQA-x1 medical imaging dataset. Paper available; pending publication in CEUR Workshop Proceedings.
QR Guard was shortlisted at a national-level hackathon for its multi-layered ML-based approach to detecting and classifying malicious QR codes.
Full-stack legal platform with a Mistral chatbot fine-tuned on Indian law data. Graph-native Neo4j backend for relationship-aware queries. Django REST API + Next.js frontend.
Automated college seat allocation portal with student preference handling, matrix-based allocation logic, and multi-round waitlist processing. Deployed on PythonAnywhere (now decommissioned).
End-to-end QR threat scanner that traces redirect chains and classifies URLs as safe, suspicious, or malicious using heuristic rules and a trained ML classifier.
Kaggle competition entry predicting market returns for Hull Tactical Asset Management. Rigorous time-series discipline throughout feature engineering and validation.
Mobile-first app that measures heart rate and SpO2 in real time using the phone's rear camera and torch via Photoplethysmography — no external hardware required.
A Linux distribution built from scratch on Debian 12 — with a fully custom display server, desktop environment, terminal emulator, and app store, all written in C.
Kappa-style streaming pipeline that ingests ~10,000 network events/second via Kafka, classifies 22 attack types across 4 categories with an ensemble ML model, and surfaces live insights on a Flask dashboard — all under 2 s end-to-end latency.
Multi-modal CLI tool comparing PDF assignments across text, code (AST), visuals (perceptual hashing), and structure — fully offline, no external API calls.
IoT crash-detection using Raspberry Pi, sensors, GPS, and GSM. Dispatches emergency alerts on collision with SOS button and environmental monitoring. IFSP collaboration.
Platform for reporting and reclaiming lost items with auth, image-based posting, and a claim workflow. Vue.js frontend with Spring Boot REST API and MongoDB.
Detects routing loops via TTL variation and packet timing analysis. Real-time Flask dashboard with configurable alerts and anomaly visualisation.
Segmentation model on the Inria aerial dataset with CLAHE preprocessing. Generates pixel-level change detection maps from before/after imagery.
Compared SGD, Momentum, Adam, RMSProp, Adagrad, Adadelta, LBFGS, and Photon on Rosenbrock, Rastrigin, Ackley, and Himmelblau benchmarks.
Static website for the Computer Science & Engineering department. Built with clean HTML, CSS, and JavaScript.
Offline competitive programming exam platform with Python & Java code execution, public/private/custom test cases, automatic time & space complexity analysis, and a professor admin panel.
Production deployment of the CP exam platform for M.Tech lab exams — runs as a systemd service on 10.6.6.1, serves 40 students over LAN with internet blocked via UFW, no external internet required.
Centralized controller for 40 Ubuntu lab machines — toggle internet per machine via UFW over SSH, manage users, distribute files, and monitor CPU/RAM/USB status from a single dashboard.
Native GTK4 + libadwaita systemd service manager for GNOME/Ubuntu. Filter, search, start/stop/restart services, view live journalctl logs — all in a non-blocking threaded UI that appears in GNOME Activities.
ESP32-based smart security system combining motion, sound, light, and ultrasonic sensors with MQTT streaming to a cloud backend. Calculates threat scores, detects anomalies, and visualises intrusion heatmaps on a live dashboard.
Open for research collaborations, ML conversations, or just a good chat about systems and language models. Currently interning as an AI Software Engineer at DriveThruData LLP — I still respond to meaningful outreach. Best reached by email.