Product-minded technology enthusiast with a B.Tech in Computer Science and Engineering (AI & ML) from VIT Chennai. I enjoy using AI-assisted development to turn messy, everyday problems into useful products and experiments.
I am drawn to the full journey: understanding a user problem, shaping the product idea, building a practical first version, and improving it through feedback. I explore applied AI, automation, and scalable product foundations - with an equal interest in technology, business context, and product management.
|
What I enjoy building
|
Open to
|
| Domain | Proficiency | Details |
|---|---|---|
| Generative AI | Hands-on | LangChain, Llama, Gemini, agent routing, prompt-driven workflows |
| Retrieval-Augmented Generation | Hands-on | Pinecone, embeddings, document-grounded search, PDF workflows |
| Natural Language Processing | Hands-on | spaCy, NLTK, named-entity recognition, regex-assisted extraction |
| Computer Vision & OCR | Hands-on | OpenCV, Tesseract OCR, image preprocessing, validation pipelines |
| Machine Learning | Exploring | TensorFlow, PyTorch, YOLO, model experimentation in notebooks |
BizGuide - AI Compliance Assistant
AI-powered multi-agent RAG platform that helps Indian businesses navigate legal, tax, and business-compliance questions through specialized, document-grounded guidance.
| Area | Details |
|---|---|
| Stack | React 19, Vite, FastAPI, LangChain, Llama 3.3 70B, Pinecone, Gemini embeddings |
| Scale | Multi-agent query routing across Legal, Tax, and General assistance flows |
| Performance | Semantic retrieval designed to ground responses in uploaded and indexed documentation |
| Security | Business-profile context, conversation history, and controlled PDF upload workflows |
| Impact | Streamlines access to complex compliance information for business users |
| Repository | View GitHub profile |
I explored a responsive React experience with a FastAPI service layer, routing requests to relevant specialist agents and using vector search to keep answers anchored in source material.
Handovr - Verified Phone Marketplace
Full-stack marketplace for buying and selling verified second-hand phones across Delhi NCR, designed around trust, device transparency, and an end-to-end handover flow.
| Area | Details |
|---|---|
| Stack | Next.js, React, Supabase |
| Scale | Seller-buyer discovery, listings, matching, and transaction workflow |
| Performance | Server-rendered product experience with responsive, streamlined listing flows |
| Security | Escrow-payment design and a 16-point device-verification workflow |
| Impact | Creates stronger buyer and seller confidence in pre-owned-device transactions |
| Repository | View GitHub profile |
The project explores how listing discovery, verification, payment safeguards, and handover can work together as one trustworthy product journey.
FinAssist - Virtual Finance Assistant
Finance dashboard that brings market data, calculators, and aggregated news into one focused interface for everyday financial exploration.
| Area | Details |
|---|---|
| Stack | Python, Flask, JavaScript, RSS integrations |
| Scale | 15+ integrated utilities, including stock, NAV, currency, commodity, and calculator tools |
| Performance | Parallel RSS fetching for multi-source market-news aggregation |
| Security | Server-side integration layer for external data services |
| Impact | Consolidates common finance utilities into a single accessible dashboard |
| Repository | View GitHub profile |
Built as a practical product experiment with SIP, EMI, FD, and Zakat calculators alongside live market-oriented utilities.
Jun 2024 - Jul 2024 Remote
Built an AI-powered invoice-extraction microservice for enterprise procurement workflows, turning unstructured invoice images into validated business data.
- Designed a four-stage pipeline: image preprocessing, OCR, NER extraction, and validation.
- Combined regex patterns with custom NER models to capture invoice numbers, vendors, dates, and amounts.
- Delivered 84% extraction accuracy across diverse invoice formats with approximately 4-second document processing.
Python Flask OpenCV Tesseract OCR spaCy REST APIs
Aug 2023 - Dec 2023 Noida, India
Developed a CRM and lead-management product with a focus on operational clarity, real-time collaboration, and efficient sales workflows.
- Built a drag-and-drop pipeline that supported management of 500+ leads.
- Implemented secure multi-user authentication, real-time analytics, meeting scheduling, filtering, and CSV export.
- Delivered a responsive interface with a 95+ Lighthouse score and workflow improvements estimated at 40%.
React Supabase PostgreSQL Tailwind CSS JavaScript
| Recognition | Details |
|---|---|
| Academic achievement | Secured a Top 3 position in school academics for Classes 10 and 12 |
| Student leadership | School Captain and School Council Committee member for three consecutive years |
| Community contribution | Operations Team Member, Business Club - coordinated technology events and workshops |
learning:
- Product management, business thinking, and user discovery
- AI-assisted development and agentic AI workflows
building:
- Useful product experiments with scalable foundations
exploring:
- Applied AI, automation, and technology-enabled business ideas
open_to:
- Product-minded teams, early-stage ideas, and collaborations
- Learning opportunities at the intersection of tech and business