Muhammad Kamran · Lahore, Pakistan
I build web applications — and the AI that goes inside them
Muhammad Kamran is a software developer in Lahore, Pakistan, with 2 years of experience building modern web applications and AI-enabled products. A strong full-stack foundation in React, Next.js, TypeScript, Laravel/PHP, REST APIs and SQL, with hands-on Python for applied AI and computer-vision work. One person, the whole stack — from the first component to the API behind it to the model it calls.
- Name
- Muhammad Kamran
- Role
- Software Developer — AI/ML & Full Stack
- Core stack
- React · Next.js · TypeScript · Laravel · Python
- Base
- Lahore, Pakistan — clients worldwide
- Experience
- 2+ years · 20+ projects
- Education
- BS Information Technology, University of Education, Lahore
- Certification
- Diploma in Machine Learning & Deep Learning — NAVTTC / CORVIT
- Languages
- English (fluent) · Urdu (native)
01 / The short version
Most products are built by a frontend team, wired up by a backend team, and handed to someone else for the AI part. Every handoff loses something.
I work across all of it. At Cross Media Sol I develop and maintain production applications in React, Next.js, TypeScript and Tailwind CSS, build the REST APIs behind them in Laravel/PHP and Node.js, and manage state with Redux and React Query. I implemented ISR with on-demand revalidation so dashboard changes reach live pages instantly, took a mobile PageSpeed score from about 55 to about 90, and put technical and programmatic SEO across production pages — on OnlineToolPot, that became an architecture for more than 130 tool pages.
The AI side is hands-on, not theoretical. On Resumaic I shipped AI-powered resume and cover-letter generation alongside a Node.js microservice for ATS scoring, with Google OAuth and Stripe subscriptions. In applied computer vision I developed experimental workflows for real-estate photography — dataset preparation, scene classification, image comparison, mask-based analysis and image-quality signals such as color, exposure, similarity and HDR differences — using Python, OpenCV, PyTorch, NumPy, Pandas, Jupyter and Google Colab to support AI-assisted quality control.
I started with a five-month frontend internship at Social Swirl, building reusable React and TypeScript components for social-media management tools. Behind that: a BS in Information Technology at the University of Education, Lahore, and a diploma in Machine Learning and Deep Learning from NAVTTC and CORVIT. I work in English (fluent) and Urdu (native).
02 / Experience
Jan 2025 — Present
Cross Media Sol
Full Stack Developer
- Developed and maintained production web applications with React, Next.js, TypeScript and Tailwind CSS — reusable interfaces, with frontend workflows integrated into backend services.
- Built and integrated REST APIs with Laravel/PHP and Node.js to support application data flows and full-stack product functionality.
- Implemented Incremental Static Regeneration with on-demand revalidation so admin and dashboard updates appear on live pages without full site rebuilds.
- Improved mobile PageSpeed from roughly 55 to 90 through dependency cleanup, dynamic component loading and font optimization; implemented technical and programmatic SEO across production pages.
Aug 2024 — Dec 2024
Social Swirl
Frontend Developer Intern
- Built reusable React and TypeScript UI components and responsive layouts for social-media management tools during a five-month frontend internship.
03 / Education
2023 — 2026
BS Information Technology
University of Education, Lahore
Sep — Nov 2025
Diploma in Machine Learning & Deep Learning
NAVTTC — CORVIT
2020 — 2022
ICS — Intermediate in Computer Science
BISE Lahore
2018 — 2020
Matriculation (Science)
BISE Lahore
04 / Toolbox
Development
- Next.js
- React
- TypeScript
- JavaScript
- Python
- PHP
- SQL
- Tailwind CSS
- Redux
- React Query
- Laravel
- Node.js
- REST APIs
- Authentication APIs
- MySQL
- HTML5
- CSS3
- WordPress
- Elementor Pro
- WooCommerce
- JetEngine
SEO & Performance
- On-Page SEO
- Programmatic SEO
- Dynamic Metadata
- Canonical Tags
- Structured Data
- XML Sitemaps
- Core Web Vitals (LCP, INP, CLS)
- ISR
- On-Demand Revalidation
- Technical SEO
- Off-Page SEO
- Local SEO
- Keyword Research
- Schema Markup
- AEO
- GEO
- AI Overviews
Tools & Infrastructure
- Vercel
- ISR & Revalidation
- Git & GitHub
- Stripe
- Jupyter
- Google Colab
- Linux
- VPS Hosting
- DNS
- SSL/TLS
- WAF
- Cloud Hosting
- Backups
- Server Hardening
AI / ML & Computer Vision
- Machine Learning
- Deep Learning
- Computer Vision
- OpenCV
- PyTorch
- Image Processing
- Image Segmentation
- Image Quality Analysis
- Scene Classification
- Image Similarity Analysis
- HDR / Image Enhancement
- Dataset Preparation
- Data Preprocessing
- Model Evaluation
- LLM / API Integration
- Prompt Engineering
- AI Pipelines
- NumPy
- Pandas
05 / Questions
Who is Muhammad Kamran?
Muhammad Kamran is a software developer based in Lahore, Pakistan, working across AI/ML and the full stack. He builds web applications and AI-enabled products with React, Next.js, TypeScript and Laravel, and uses Python, OpenCV and PyTorch for applied AI and computer-vision work. He publishes his work as KamranDev.
Where is Muhammad Kamran based, and does he work remotely?
He is based in Lahore, Punjab, Pakistan, and works remotely with clients worldwide in English, on Pakistan Standard Time (UTC+5). The portfolio includes work for businesses in Uzbekistan, United Kingdom, United States, Venezuela, New Zealand.
What technologies does he work with?
Frontend: React, Next.js, TypeScript, Tailwind CSS, Redux and React Query. Backend: Laravel/PHP, Node.js, REST APIs and MySQL. AI/ML: Python, OpenCV, PyTorch, NumPy, Pandas and LLM/API integration. For content sites: WordPress, Elementor Pro and WooCommerce.
What AI and machine-learning work has he done?
On Resumaic he built AI resume and cover-letter generation alongside a Node.js microservice for ATS scoring and resume parsing. On a real-estate image QC project he built experimental computer-vision workflows in Python, OpenCV and PyTorch — scene classification, image comparison and image-quality signals. He holds a Diploma in Machine Learning & Deep Learning from NAVTTC — CORVIT.
Does he take freelance work?
Yes. Besides his role at Cross Media Sol, he works as a freelance developer — the client websites in his portfolio are freelance projects, for businesses in Uzbekistan, United Kingdom, United States, Venezuela, New Zealand.
How much experience does he have?
2 years. He is currently a Full Stack Developer at Cross Media Sol (Jan 2025 to Present), after a five-month frontend developer intern role at Social Swirl (Aug 2024 to Dec 2024).
What is his education?
He studied for a BS in Information Technology at the University of Education, Lahore (2023–2026), and holds a Diploma in Machine Learning & Deep Learning from NAVTTC — CORVIT (September–November 2025).
What services does he offer?
Web Application Development, AI Solutions, Website Development, SEO & AI Search, WordPress Development, Website Design, Website Performance, E-Commerce & WooCommerce and Website Redesign.
How do I contact Muhammad Kamran?
Email info@kamrandev.com, message or call +92 320 6899684 on WhatsApp, or use the contact form at kamrandev.com/contact. He usually replies within a day.
See how this shows up in practice — browse my work or read my process.
Next Step
Have an idea worth building?
Let's turn it into something people remember — and search engines can understand.
Usually replies within a day · Lahore, Pakistan · working worldwide
View my work