AI AND ML ENGINEER
building products that dont suck
I’m an AI & Machine Learning Engineer who enjoys building systems where models meet real users. My work involves developing ML backed applications, experimenting with retrieval augmented generation, and designing pipelines that turn unstructured data into usable intelligence. While actively building products, I’m deepening my understanding of machine learning concepts such as data preprocessing, model behavior, and evaluation to create more reliable and scalable AI systems. I’m driven by curiosity, hands-on experimentation, and the goal of translating ML theory into practical impact
Gen-Z Tech Newsletter Editionaly is a Gen-Z–focused newsletter that delivers tech news, internet trends, and stories in a comic-style format. Instead of long articles or boring headlines, information is presented visually with humor, memes, and bite-sized storytelling—making news easy to consume and fun to read. The goal is to keep readers informed while matching the fast, scroll-friendly habits of modern audiences.
Faith-Centered AI Chatbot Using RAG SiratAI is a lightweight Islamic chatbot built using retrieval-augmented generation (RAG). It provides accurate, grounded answers by retrieving relevant Quranic verses, Hadiths, and Seerah sources before generating responses. The system is designed to prioritize authenticity, transparency, and respectful handling of faith-based queries.
HatchUp is an actively evolving project that I am currently building and iterating on, with a strong focus on experimentation, user feedback, and continuous improvement. The work so far involves designing core systems, refining workflows, and testing early versions with students and founders to understand real-world needs. The project has received early validation in the form of a ₹10,000 grant, which is being used to further develop and strengthen the foundation as the product continues to take shape.
AI Document Intelligence Agent with Confidence Scoring Documind is an AI-powered document parsing system built with Streamlit. It processes PDFs and images—including handwritten notes—to automatically detect document types such as invoices, receipts, prescriptions, resumes, and more. Using OCR combined with LLMs, it extracts structured JSON outputs and assigns confidence scores at both field and document levels to ensure reliability.
QR-Based Food Ordering and Delivery System This project automates the drive-thru and parking-based food ordering process using QR codes. Customers place orders by scanning a QR code, which are instantly routed to the kitchen and assigned to waiters with precise parking location details. The system reduces wait times, improves coordination, and minimizes operational confusion in high-traffic food outlets.
AI-Powered Legal Document Simplification Platform DocEase is a Streamlit-based web application that simplifies complex legal documents into easy-to-understand summaries. Users can upload legal PDFs, translate content into regional languages, and generate concise explanations using large language models. The app also supports audio playback of the simplified text, making legal information more accessible to non-technical users.
Secure Real-Time Chat Application with AI Assistance Chatrixx is a Python-based real-time chat application designed for secure communication over a local network. It implements end-to-end encryption to ensure message privacy and integrates an AI assistant powered by Google Gemini. Users can invoke the AI directly within conversations using the /ai command, enabling instant intelligent responses without leaving the chat environment.
At Loop, I worked as a Generative AI Engineer where I designed and developed a Retrieval-Augmented Generation (RAG) application to enable accurate, context-aware AI responses. The system ingested and indexed documents, generated vector embeddings, and performed semantic retrieval to fetch relevant context before response generation. My work focused on improving answer reliability, reducing hallucinations, and optimizing retrieval pipelines for real world usage. This role involved hands-on experimentation with LLMs, prompt design, and end-to-end AI system integration.
I worked as a freelance graphic designer, creating posters, thumbnails, and social media creatives for multiple clients across different domains. Beyond design execution, this role helped me develop strong client relationship skills understanding requirements, handling feedback, managing revisions, and delivering work on time. I was also hired by an esports team, where I collaborated closely with stakeholders to produce visuals aligned with their brand and audience. This experience strengthened my communication, professionalism, and ability to translate client needs into effective visual outcomes.
No research papers yet.