Business websites
Modern marketing and ordering sites
Over the past few years of hands-on software engineering, I have designed and shipped end-to-end systems under the HysMTech brand — spanning modern business websites, autonomous AI agents, multi-step automation workflows, and production-grade APIs. Every project reflects the same standard: clean architecture, thorough documentation, and measurable business impact. From conversion-focused web apps to intelligent agents that reason, act, and integrate with your existing stack — I build systems that work reliably long after launch.
Focus Areas
What I build
AI employees
Role-based agents in your workflows
Agentic systems
Multi-step reason → act → verify loops
Case Studies
Projects, in detail.
Client website
A One Juice
High-energy product site for a juice & ice cream business with clear menu hierarchy and order CTAs.
View live
Client website
Karachi Hotel Sahiwal
Restaurant brand site featuring signature dishes, menu navigation, and order-focused UX.
View live
AI employee
MailVox
An AI employee that monitors Gmail, reads incoming email, and automatically creates replies according to a company policy PDF you attach. The agent follows that policy and, after your approval, sends the reply.
View live
AI agent (personal AI employee)
Kropost
Assign a topic → agent researches → generates an image and a professional draft → waits for your approval → publishes to LinkedIn automatically.
Case study only — no public URL
Autonomous software engineering agent
Remobug
Transforms plain-English tasks into working code edits across your repository. Handles bug fixing by retrieving relevant files, planning precise edits, and executing test suites. Built-in self-correction loop parses test failures and iterates in a secure isolated environment until success. Fully deployed to accelerate development workflows.
View live
SaaS product (personal build)
QueueNova
Digital queue management for hospitals, banks, and any counter-based business. Customers book their token from a mobile phone, watch their position live, and get notified when 2–3 people are left in front — so they walk in just in time instead of waiting in line. One system, any industry where people queue: less rush, no wasted hours, fair turns.
Launching soon — final stage before public releaseSkills
The toolkit behind the systems.
Languages & APIs
- Python
- FastAPI
Retrieval
- RAG
- Agentic RAG
- CRAG
Agents & Orchestration
- LangChain
- LangGraph
Delivery
- Docker
- Azure (deployment)
Credentials
Verified certificates.
Multi-Agent Systems with LangGraph
Coursera
Advanced RAG with Vector Databases and Retrievers
Coursera
Python Object Oriented Programming
Coursera