Muhammad
Zohaib Zahid
I build the intelligence layer — from scalable full-stack systems to LLM pipelines and agentic AI. Production-grade, always.
I don't just build features.
I engineer systems.
Five years building software that actually ships — from responsive React frontends to backend architectures handling real production load. My work sits at the intersection of full-stack engineering and applied AI.
I care about the code that comes after mine: typed, documented, built for the engineer who inherits it. Whether it's an LLM workflow system, a 3D virtual store CMS, or a fintech compliance platform — I treat every system as if it has to outlive me.
Currently at Directed Trust Co. in Phoenix (remote), building a secure IRA platform. Previously at Geekybugs, where I led frontend architecture and delivered production AI features.
Projects
Real systems, shipped to production. Click any card to explore full highlights.
Walmart Virtual Store CMS
A backend CMS powering a mobile app for virtual stores, enabling custom themes, 3D product placements, multi-step approval workflows, and GCP deployments.
Emperia QuickStitch Dashboard
A dashboard for rendering and managing 3D virtual worlds via JSON config uploads. Multi-org management, per-tier bandwidth allocation, and Stripe subscriptions.
Directed IRA Platform
Secure self-directed IRA platform with OWASP-hardened backend, Alloy identity verification, CSRF/XSS protection, and scalable user-facing investment workflows.
Dark Weather — AI Weather App
Async-first weather app with an agentic AI assistant using OpenAI SDK + vector DB for persistent conversational context. Response time cut from 2–3s to under 1s.
Hai-Sensei — LLM Workflow System
AI backend that transforms meeting transcripts into structured behavioral analysis reports through a 20–27 stage prompt chain. Built for accuracy, continuity, and scale.
Career Path
Software Engineer
- →Built a secure, scalable self-directed IRA platform enabling users to manage alternative investments with regulatory compliance.
- →Improved platform security ~40% via OWASP hardening — auth, authorization, and data protection.
- →Integrated Alloy identity verification, cutting fraudulent account creation by 15%.
- →Implemented CSRF/XSS protection, Helmet headers, session management, and secure cookies.
Associate Software Engineer
- →Optimized application performance across frontend and backend — refactoring, state management, best practices.
- →Delivered pixel-perfect web apps migrated from legacy frontend views.
- →Designed and consumed REST APIs with seamless frontend integration.
- →Mentored junior interns and participated in code reviews and architecture discussions.
BS Software Engineering
Coursework: Object Oriented Programming, Data Structures & Algorithms, Database Systems, Machine Learning, Deep Learning, and Information Security.
Tech Stack
What I actually use in production — not just a buzzword list.
Frontend
Backend
AI / ML
Data
Infrastructure
Languages
The Process
Every engagement, every project — same five steps.
Understand
Clarify the real problem before touching code. Align on constraints, users, and outcomes.
Architect
Data models, API boundaries, component structure — designed before the first line.
Build
Typed, precise, tested. Clean enough for the engineer who comes next.
Integrate AI
LLM pipelines, vector stores, agentic tools — where they earn their place.
Ship
Deploy, monitor, iterate. Production is where real engineering begins.
Certifications
The Complete Agentic AI Engineering Course
Ed Donner · Udemy · 2025
LLM Engineering: Master AI, LLMs & Agents
Ed Donner · Udemy
Node.js, Express, MongoDB & More
Jonas Schmedtmann · Udemy
Web Development using ReactJS 2022
Udemy
Let's build
something real.
Open to senior engineering roles, AI projects, and interesting problems. Remote-first. Based in Lahore, PKT (UTC+5). Usually responds within 24 hours.