I started out in Electrical and Electronics Engineering, then moved into software because I wanted to build the systems that actually run on the hardware I was studying rather than just the hardware itself. That shift has shaped how I work: I care as much about why something works as whether it does.

Over 6+ years at McAfee and Cognizant, I've built and shipped production frontend and full stack systems using React, Angular, Vue, and Node.js, with a strong focus on accessibility (WCAG) and test coverage (Jest, Cypress). I'm currently completing an MSc in Advanced Software Engineering at the University of Leicester, where I'm building an AI-powered platform from the ground up as my dissertation project.

I'm actively looking for Frontend Engineer or Software Engineer roles where I can bring that mix of production experience and hands-on AI/ML curiosity.

Languages

Frontend

Backend

Testing & Quality

Cloud & Tools

Certifications: Azure Fundamentals · Azure AI Fundamentals · PG Diploma in Data Science and Business Analytics (UT Austin)

• Four distinct user roles (buyer, developer, supplier, admin) with server-side role-based access control enforced on every route, plus ownership-level checks so a developer can only act on their own plots and a supplier only on their own orders, not just role checks alone.
• AI recommendation engine built on FAISS vector search and sentence-transformer embeddings, scoring the product catalog against each buyer's questionnaire answers in a three-pass fallback, so buyers always get relevant results even with narrow filters.
• AI buyer assistant powered by Gemini, with its own retrieval step over products and FAQs, a defensive system prompt that treats all retrieved/user data as untrusted input, and regex-based prompt-injection detection before any message reaches the LLM.
• Automated end-to-end order workflow: approving a buyer's selections auto-generates a PDF summary, splits line items into separate purchase orders per supplier, and notifies every affected party in real time.
• Real-time messaging and notifications via Socket.IO, with a relationship-derived contact list, buyers, developers and suppliers can only message people they actually have an active order relationship with, not an open directory.
• AI-assisted invoice processing: suppliers upload invoices to S3, and a Gemini-backed extraction endpoint summarises each one and flags amount mismatches against the original purchase order automatically.
• Secure file handling throughout: all uploads (invoices, chat attachments, floor plans) are stored privately in S3 and only ever exposed via short-lived, time-limited signed URLs, never public links.
• Budget-aware checkout logic that tracks a plot's fixed extras allowance against both prior approved spend and the live basket total, added without touching any of the existing, already-tested pricing calculations.
• Thoroughly tested: 380+ automated test cases across four frameworks (Jest, Vitest, React-Testing-Library, pytest, Playwright), including real-database and real-HTTP integration tests, not just mocks.

A full-stack MERN application for running referendums, with fully separate Voter and Election Commission experiences.

• Built separate Voter and Election Commission dashboards with distinct permissions and views
• Implemented JWT-based authentication with role-based authorization and bcrypt password hashing
• Built a QR code scanning flow for secure voter check-in
• Delivered results visualization using Chart.js (bar/donut charts) and a custom word cloud view for standout voting options
• Exposed a public open-data REST API for referendum results (/mslr/referendums, /mslr/referendum/:id)
• Backed by MongoDB Atlas with Mongoose schemas across voters, referendums, options, votes, and audit history

Stack: React (Vite), React Router, Axios, Bootstrap, Chart.js, Node.js, Express.js, JWT, bcrypt, MongoDB Atlas, Mongoose

• food-delivery-app — React app consuming the Swiggy API, with restaurant search, filtering by rating, and detail/menu pages
• shopping-cart — React shopping cart with product search/filtering and a full cart flow
• mini-store — React storefront focused on state management with the Context API
• pagination-project — React product listing app with API-driven, client-side pagination

📫 Reach me on LinkedIn — I'm actively interviewing for Frontend and Software Engineer roles and always happy to talk shop about React, accessibility, or applied AI.