Available for new opportunities

Talha Zia.

Frontend & Full-Stack Developer

I build fast, responsive web products with Next.js and Laravel — and I'm currently learning how neural networks actually learn.

See my workGet in touch
Now
Frontend Developer at Al-Rehman
Learning
Neural Networks with PyTorch
Based in
Rawalpindi, Pakistan

01About

Clean code, real products, constant curiosity.

I'm a frontend and full-stack developer who ships real products — for travel, real estate, ERP and e-commerce.

Day to day, that means Next.js, Redux and SCSS on the frontend, with Laravel and MySQL behind it. I care about clean code, fast pages and interfaces that hold up on every screen. Alongside work, I'm studying for a BBIT degree and teaching myself AI — starting from the fundamentals of neural networks.

  1. 01

    Detail-driven UI

    Spacing, motion and every interaction state are designed on purpose, not left to chance.

  2. 02

    Fix it at the root

    Auth flows, API errors, state bugs — I chase them until they're properly solved, not patched.

  3. 03

    Mobile first, always

    Every layout is built and tested from the smallest screen up.

Toolkit

What I build with.

AI/ML tools are in the AI Journey

Frontend

  • Next.js
  • React
  • TypeScript
  • JavaScript (ES6+)
  • Redux
  • Sass / SCSS
  • HTML5 & CSS3
  • jQuery

Backend & Data

  • Laravel
  • PHP
  • Node.js
  • Express
  • REST APIs
  • MySQL
  • MongoDB

Tools & Workflow

  • Git & GitHub
  • Jira
  • Postman
  • Figma
  • Vercel
  • VS Code

02Experience

Where I've shipped.

Onsite roles across travel, real estate, ERP and full-stack products — from internship to full-time.

  1. Jan 2026 — Present

    Rawalpindi · Onsite

    Frontend Developer

    Current

    Al-Rehman·Intern → Full-time

    • Built the schedule and seat-selection flows for a production travel app, responsive across every screen size.
    • Added theme switching and multi-language support, with app state managed in Redux.
    • Refined seat-selection logic and layouts for a smoother booking experience.
    • Next.js
    • SCSS
    • Redux
    • Jira

  2. Sep 2025 — Dec 2025

    Rawalpindi · Onsite

    Full-Stack Developer

    Friends IT Solution·Intern → Full-time

    • Fixed critical authentication and session-security issues.
    • Built the Add Property feature for the MyBahria real-estate admin panel.
    • Contributed employee-leave and management modules to an ERP system (Fisdemo).
    • Resolved CRUD bugs and polished the UI/UX of admin dashboards.
    • Laravel
    • PHP
    • MySQL
    • Blade
    • JavaScript

  3. Contract

    Rawalpindi · Onsite

    Frontend & API Integration Developer

    SkilledMan & FLucom·Contract

    • Integrated backend APIs to drive dynamic, real-time UI updates.
    • Debugged API responses and error handling with Postman and cURL.
    • Worked closely with backend engineers to keep data flowing reliably.
    • JavaScript
    • jQuery
    • Laravel API
    • Postman

  4. Aug 2025

    Rawalpindi · Onsite

    Full-Stack Developer Intern

    MorosoftLabs·Internship

    • Built full-stack web applications on the MERN stack.
    • Implemented REST APIs and end-to-end CRUD features.
    • MongoDB
    • Express
    • React
    • Node.js

03Projects

Selected builds.

Web projects I've designed and developed end to end. Each one links straight to its source code.

More on GitHub

04AI Journey

Learning how machines learn.

I'm a web developer learning AI/ML from the ground up, with neural networks as my current focus. This is an honest log — what I've covered, what I'm studying now, and what comes next.

Current focus

Neural Networks

Not just calling a library — understanding how layers, weights and gradients work together, and why a model gets better with every batch.

  • Layers, weights & activations
  • Forward pass & backprop
  • MNIST classifier

Forward pass

4 → 6 → 6 → 3
InputHiddenHiddenOutput

Illustrative · hover or tap a neuron to trace its weights

ALearning path

In the order I'm tackling it

  1. 01

    Python

    The language behind every experiment I run.

    Completed
  2. 02

    PyTorch

    Tensors, autograd and the fundamentals of a training loop.

    Completed
  3. 03

    NumPy

    The array maths that every neural network layer is built on.

    Learning now
  4. 04

    Pandas

    Loading, cleaning and exploring data before it reaches a model.

    Learning now
  5. 05

    Neural Networks

    My main focus right now — understanding how networks actually learn.

    Learning now
  6. 06

    Deep Learning

    Moving beyond a single hidden layer toward deeper architectures.

    Learning now
  7. 07

    Computer Vision & NLP

    Applying what I've learned to images and text — and bringing models into web apps.

    Up next

  8. Progress

    • Completed2
    • Learning now4
    • Up next1

BConcept lab

Interactive notes from what I'm studying

The training loop

Four steps every PyTorch model repeats, batch after batch.

Step 1 of 4

Forward pass

Data flows through each layer: multiply by the weights, add a bias, apply an activation. The final layer produces the prediction.

>>> logits = model(images)

Activation functions

Pick a function, then drag the input to see its output.

f(x) = max(0, x)

Default for hidden layers — cheap to compute and keeps gradients alive for positive inputs.

CExperiments & what's next

Real repos first, then plans

Built

Planned · not started yet

  • 01

    Digit recognizer on the web

    Serve my MNIST model behind a Next.js canvas — draw a digit, see the prediction live.

    PyTorch · Next.js

  • 02

    Image classification with CNNs

    Move from fully connected layers to convolutions as my first real computer-vision step.

    Computer Vision · PyTorch

  • 03

    Text classification basics

    Tokenization, embeddings and a simple sentiment classifier to explore NLP fundamentals.

    NLP · Pandas

05Contact

Have an idea? Let's build it.

I'm open to new roles and projects. Send a message and I'll get back to you quickly.

Write to me

Based in Rawalpindi, Pakistan