About Me
I build software where data, cloud, and intelligence meet — and I'm obsessed with shipping things that actually work in production. I'm a Computer Science (Data Science) engineer at RCOEM (9.65 CGPA) and currently a Product Development Intern at Darwinbox, where I work across the full stack — React, Node.js, Express, and MongoDB — turning real product requirements into clean, scalable features. My favourite lesson so far: writing good code beats writing more code, and strong fundamentals make everything downstream easier. My range is intentional. I've shipped deep-learning and computer-vision projects (emotion, drowsiness, and sign-language detection) during my Data Science internships, earned the AWS Certified Cloud Practitioner and Oracle AI Vector Search Professional credentials, and built hands-on experience with the things AI products are actually made of today: vector search, embeddings, RAG pipelines, and cloud-native deployment on AWS. That combination — software engineering + DevOps + applied ML — is exactly where I like to operate. Beyond code, I led publicity for DASCA (Data Science Association) and contributed to PR & outreach for the GeeksforGeeks RCOEM Chapter — roles that taught me to communicate technical work to non-technical audiences and ship under deadlines with a team. What I'm looking for: SDE, DevOps, ML/AI, or Data Engineering roles where I can build products at scale, learn from strong engineers, and keep closing the gap between "interesting model" and "reliable system." Let's connect — I'm always up for a conversation about cloud, ML, or building things that last. 🛠 Core stack: Python · AWS · Node.js · MongoDB · SQL · Docker · PyTorch · RAG/Vector Search
Skills & Technologies
Work Experience
Product Development Intern
Built Darwinbox’s first LLM-powered AI form automation platform, reducing HR form creation effort by 80%. Created AI agent–based automation tools, improving operational efficiency by 60%. Designed and developed a scalable Payroll Operations Platform for managing payroll workflows.
Projects
ClassVision: Advanced Classroom Activity Recognition
Developed a pose-based YOLO + MediaPipe + LSTM model for multi person activity recognition, improving accuracy by 40%. Built a real-time sequence modeling pipeline for behaviour classification.
Oncology Document Analyser System
Built an NLP system to classify cancer records with 90%+ accuracy. Developed an LSTM + Word2Vec sequential model for performance on unstructured medical text classification.
One-Shot: XAI-Powered Exam Prep Platform
Built an AI platform for summaries, quizzes, and adaptive tests, reducing effort by 60%. Implemented RAG with LangChain + Pinecone, boosting preparation efficiency by 50%.
Education
B - Tech
Data Science