Hi, I'm
MISAL
AZEEM
AI-First Full Stack Engineer
Building production-grade web applications and scalable GenAI systems. Led RLHF pipelines for Google DeepMind's Gemini models. 5+ years shipping code that matters.
Building at the
intersection of AI & Web
I'm an AI-First Software Engineer and Full Stack Architect based in Islamabad, Pakistan, with over 5 years of experience building production-grade web applications and scalable GenAI systems.
My work spans leading RLHF and QA pipelines for Google DeepMind's Gemini models, driving 50% AWS infrastructure cost reductions, and delivering enterprise AI proof-of-concepts from architecture to deployment. I bridge advanced AI capabilities with production-ready code.
I thrive in cross-functional teams, having mentored 12+ engineers and collaborated with researchers to refine evaluation guidelines for cutting-edge AI models.
Where I've made impact
Frontend AI Evaluator
Revelo
- Engineered pixel-perfect "golden solution" UI components from Figma designs using HTML, CSS, and JS for AI model training ground-truth datasets
- Audited AI-generated code against complex behavioral prompts, identifying architectural flaws to improve model accuracy
Senior AI Software Engineer & QA Lead
Invisible Technologies (Meridial AI)
- Served as IC and QA Lead for Google DeepMind training initiatives with high-fidelity code annotations
- Evaluated AI agent trajectories in complex scenarios with deterministic scoring for benchmarking
- Contributed to Kotlin open-source repos: triaging issues, developing patches, and engineering test suites
AI First Software Engineer
Crossover
- Led migration of Khoros community platform to AWS Graviton, reducing EC2 costs by 50%
- Implemented cost optimization across AWS RDS, OpenSearch, and Kinesis for quarterly savings
Full Stack Engineer
Mutekra
- Developed three enterprise-grade AI/GenAI POC apps using Python, FastAPI, TypeScript, React, LangChain, LangGraph, and Qdrant
- Designed Muswarm: swarm-based AI agent visualization with LangGraph and Model Context Protocol orchestration
- Engineered Musupportra: agentic chatbot using LangChain for autonomous Tier 1 ISP support
POD Lead
Turing
- Led cross-functional pod for code quality and RLHF on Google DeepMind's Gemini models
- Optimized data pipelines across Python, Java, and JavaScript for AI training acceleration
- Mentored 12+ engineers on high-quality data generation and evaluation guidelines
Front End Team Lead
AAk Tele-Science
- Reduced website load times from 8s to 2s (75% improvement) via performance optimization and Tailwind CSS migration
- Integrated Redux state management and built a full-featured dashboard with subscription management
Technical Lead
RLAB Group
- Delivered CMS solutions achieving 95% client satisfaction and 20% boost in project efficiency
- Implemented QA pipelines reducing post-launch issues by 95%
My technical arsenal
Languages
Frontend
Backend
Databases
Cloud & DevOps
AI & ML
Featured work
A selection of projects from my GitHub. These range from full-stack applications to frontend showcases, demonstrating breadth across React, Next.js, Rails, and more.
Blog Application
Full stack blog platform built with Next.js and TypeScript. Features server-side rendering, dynamic routing, and a modern UI.
Webstore E-Commerce
Complete e-commerce platform with React/TypeScript frontend and Node.js/Express/MongoDB backend. Full product catalog with shopping cart.
Space Travelers Hub
React/Redux application fetching from the SpaceX API. Users can join missions and reserve rockets with persistent state management.
Task Management App
Full stack task management with JWT authentication, user management, and a clean UI. Separate frontend and backend architecture.
Concert Event Platform
Responsive concert event website showcasing CSS Grid/Flexbox mastery, mobile-first design, and vanilla JavaScript interactions.
Let's build
something great
I'm always open to discussing new opportunities, interesting projects, or just having a conversation about AI and technology.