Loading Portfolio
Product · AI · Engineering
|
I ship production-ready apps end-to-end — built with LLMs, agentic workflows, and AI coding tools like Claude Code & Cursor.
The mind behind the machine
Product & AI Engineer who ships production-ready applications end-to-end — from product discovery and system design to deployment and iteration. Specialized in LLM systems, agentic workflows, and scalable backend architectures using LangGraph, FastAPI, and microservices, with deep experience in RAG pipelines, computer vision, and ML classification systems.
Day-to-day I work hands-on with agentic coding tools — Claude Code, Cursor, GitHub Copilot, Windsurf — to compress build cycles and ship faster without compromising quality. Proven ability to design memory-efficient AI agents, human-in-the-loop workflows, and scalable architectures that survive real production load.
The Scholar's Library
The Exhibition Hall
AI / LLM🏠AI-Powered Interior Design Mobile App
AI / LLM📸AI Photo Enhancement & Virtual Try-On Platform
AI / LLM🏥AI-Powered Healthcare SaaS Platform (UAE Clinics)
AI / LLM🤖AI-Powered Twitter/X Automation Platform
ML / CV🫀ECG Intelligence & Doctor Verification Platform
Mobile🌙AI-Powered Sleep Tracking & Wellness Mobile App
Mobile🍵Anonymous Social Platform with Community Verdict System
The Journey
Continuous Learning
Professional Overview
AI Engineer | LLM Systems | Backend & Agentic Architect
imosmanwaris.tech@gmail.com | +923200787777 | GitHub | LinkedIn
AI Engineer specializing in LLM systems, agentic workflows, and scalable backend architectures. Experienced in designing production-grade AI systems using LangGraph, FastAPI, and microservices. Strong background in RAG pipelines, AI automation platforms, computer vision, and ML classification systems.
GPA: 3.06
Programming: Python, FastAPI, Flask, Django, JavaScript/TypeScript
AI/ML: LangChain, LangGraph, RAG, Transformers, OpenCV, PyTorch, Scikit-Learn
Cloud & DevOps: Docker, Kubernetes, AWS, Google Cloud, CI/CD, MongoDB, PostgreSQL, Redis
What People Say
Usman brings exceptional technical depth to every project. His ability to architect complex AI systems with LangGraph while maintaining clean, production-ready code is remarkable.
Notes on Products, AI-Powered Products & Agentic AI
Most AI projects die between the notebook and the user. Here's how I think about scoping, latency budgets, and fallback states so a model actually becomes a product.
6 min readStreaming, optimistic UI, graceful degradation, and trust signals — the UX patterns that turn an LLM call into a product users actually rely on.
8 min readWhat it actually takes to run agents in prod — memory budgets, retry strategies, human-in-the-loop checkpoints, and observability you can debug at 2am.
9 min readA breakdown of the components every agent needs — tool registries, scratchpads, working memory vs. long-term memory, and how to keep the planning loop from running away.
7 min readHow I use Claude Code and Cursor day-to-day to compress build cycles — from spec to PR — without dropping quality. Hooks, MCP, slash commands, and the prompts that actually move the needle.
10 min readWalkthrough of taking an AI-powered product from a one-pager to a deployed app in a week — product spec, system design, agentic build, deploy, and the things I'd do differently.
11 min readApp design concepts brought to life
A creative app design concept showcasing modern UI/UX patterns, smooth interactions, and thoughtful visual design.
Open Source Contributions
Let's build something extraordinary together
