I build software thatships fast.
Four years deep in two hard things: AI systems that do real work, and blockchain platforms that move real money. I have shipped three RWA tokenization platforms across Ethereum, BNB Chain and Algorand, plus AI agents, RAG pipelines and medical CRMs running automation in production.
Bring me the problem everyone else called complicated. I scope it, build it, deploy it, and stay until it works.
What I build
Four disciplines. One engineer.
AI Agents and RAG Systems
I build AI that does the work, not AI that demos well. Agents that take actions, pipelines that answer from your own data with citations you can trust.
- Multi step agents with tools, memory and guardrails
- RAG pipelines with vector search and source citations
- Workflow automation that removes manual hours
- LLM cost and latency tuned for production traffic
RWA Tokenization and Web3
Three real world asset platforms shipped across three chains. Contracts, compliance and money flow handled together, because on chain mistakes are expensive.
- Fractional ownership of real assets via token standards
- Investor onboarding, KYC gates and admin workflows
- Contract development, testing and mainnet deployment
- Fiat and on chain payment rails in one system
Medical CRMs and Business Systems
Complex internal systems with real users, real records and real compliance pressure, with AI automation wired in where it saves the most time.
- Patient and practice CRMs with role based access
- AI automation for intake, notes and routing
- Billing, scheduling and reporting dashboards
- Integrations with the tools the team already uses
Full Stack Product Engineering
The layer that turns a model or a contract into something people can actually use. Fast frontends, solid APIs, deployment that does not wake you up at night.
- Next.js and MERN applications built to scale
- API design, auth, queues and background jobs
- Cloud deployment with CI/CD and staging
- Performance, SEO and analytics done properly
Real world assets
Three RWA platforms. Three different chains.
Tokenizing real assets means the contract, the compliance and the money flow all have to be right. I have done it end to end on three networks, each with its own tooling and its own constraints.
Property tokenization where investors buy fractional ownership through ERC-20 tokens, with campaign pages, investor dashboards and full admin control over issuance.
A lower fee tokenization platform where the economics only worked on cheap transactions. Built the contracts, the distribution logic and the investor flow around that constraint.
Tokenization on Algorand where instant finality and built in asset controls mattered more than ecosystem size. Different toolchain, same discipline from contract to dashboard.
Selected work
Products I’ve shipped.
How I work
A tight loop, idea to production.
Understand
I dig into the real problem, the users and the constraints, then come back with a plan and a straight estimate.
Architect
Data model, contracts, agent design, infrastructure. The decisions that are expensive to change get made first.
Build
Tight weekly sprints with working software you can click. You always know where the project stands.
Ship and hold
Deploy, monitor, harden and keep improving. I stay past launch, because launch is when the real feedback arrives.
Experience
Four years, shipping the hard parts.
Full Stack and AI Engineer
- Built three RWA tokenization platforms on Ethereum, BNB Chain and Algorand
- Shipped an embeddable RAG chat widget and AI agents into live client products
- Delivered LightNX, a full defence procurement platform
- IAMDivinity billing platform joining fiat and on chain payments
- Built TOWINTEL, a live dispatch platform for a real towing company
- StablePal, a Solana payments platform for freelancers and gig workers
- Owned deployment, CI/CD and staging pipelines end to end
Freelance Engineer
- AI automation builds for small teams and founders
- Smart contract development and dApp integration
- Practice management CRMs for a chiropractic clinic and a medical billing office
- Shopify and storefront builds for UAE and European clients
Computer Science
- Studying computer science alongside full time engineering work
- Currently going deeper on agent orchestration with LangGraph
Stack
What I reach for when it has to work.
Testimonials
Clients keep coming back.
Umair replaced what we thought needed a whole engineering team. He shipped in weeks what had stalled for months.
He took our RWA idea from a rough deck to a working platform on chain, and explained every decision along the way.
The AI automation he built saves our staff hours every single week. It just works, quietly, in the background.
Contact
Tell me the problem. I will tell you how I would build it.
Hiring for a role or scoping a build, either way you get a straight answer within 24 hours. No sales call needed first.