Building Scalable & AI-powered Full-Stack Software Applications

Based in Pakistan, I am an AI Software Engineer passionate about creating dynamic, intuitive, user-centeric and AI-powered software applications.

Download Resume

About Me

Specialized in the Full-Stack and Agentic AI Development
with practical expertise in building secure REST APIs, Redis caching layers, Docker containerization, and advanced RAG pipelines. Passionate about bridging the gap between scalable backend architecture, intuitive interfaces, and AI.

Skills & Tools

Languages & Frontend

JavaScript
TypeScript
HTML
CSS
Tailwind CSS
React
Redux Toolkit
Next.js

Backend & Databases

Node.js
Express.js
MongoDB
PostgreSQL
Redis
Docker

AI & Tools

LLM
LangChain
RAG
Git
GitHub
Postman
Linux
Figma

Experience

Backend AI Engineer Intern: FlyRank AI; - Developed production-grade REST APIs containerized with Docker, integrating LLMs and leveraging PostgreSQL alongside vector databases and embeddings for multimodal semantic matching. - Implemented durable background workers and scheduled queues via Redis and BullMQ to guarantee idempotent processing and high-performance caching.

Projects

DocuMind: Agentic & Corrective RAG System for Document Interaction

DocuMind is an Agentic and Corrective RAG platform engineered to deliver verifiable, context-grounded answers from complex PDFs. It features layout-aware ingestion that preserves tabular data, parallel multi-query hybrid search (BM25 + Dense Vector), and Jina reranking. The architecture integrates a self-correcting Grader Agent, autonomous Tavily web search fallback, and a split-screen workspace where users can read and annotate documents alongside a chat interface with clickable page-level citations and streaming audio playback. Evaluated via an automated LLM-as-a-judge pipeline in LangSmith, the system achieved 91.67% overall accuracy across a 60-example test suite.

category: AI Engineering, Web Development

tags: Web Development, Full Stack Development, AI Engineering

Tech Stack: TypeScript, Next.js, MongoDB Atlas, LangChain, Groq LPUs, Hugging Face, Jina AI, Tavily, Supabase, Tailwind CSS

Links

Linkest: Intelligent URL Shortener

Linkest is a high-performance URL shortener engineered for speed, security, and real-time insights. Utilizing Redis caching to reduce database read load by 80%, delivering sub-15ms redirection through an efficient cache-aside pattern. It features AI-powered link previews using real-time content summarization, automated QR code generation, and custom aliases. To ensure enterprise-grade security and monetization, the platform incorporates password protection, auto-expiring links, Stripe subscription tiers, and interactive analytics for tracking link engagement.

category: Web Development

tags: Web Development, Full Stack Development, AI Engineering

Tech Stack: TypeScript, Next.js, Redis (Upstash), MongoDB Atlas, Supabase, Clerk, Stripe, Tailwind CSS, MUI

Links

Natours: Tour Booking Application

A comprehensive tour booking solution focused on secure user workflows and seamless transaction handling. It features a complete authentication system with JWT-based session management, password resets, and role-based access control for administrative tasks. Users can explore detailed itineraries with geolocation data, submit ratings, and securely book adventures via an integrated Stripe payment gateway. The platform is architected as a robust RESTful API, ensuring efficient data management for tours, reviews, and bookings.