Syed Asjad Abbas — AI & Full-Stack Engineer

👋 Hi, my name is

SYED ASJAD ABBAS

Organization

Founder @ Zeploy Tech

LinkedIn GitHub Zeploy Tech

AI/ML ENGINEER

& Full-Stack Developer

Focused on backend systems, AI workflows, and production-focused applications that need to handle real users, real data, and real operational constraints.

Portfolio

EXPLORE WORK

10+ Projects Shipped

3 Production Deployments

AI & Backend Core Focus

Zeploy Tech Founder

About

Engineering at the edge of AI

Final-year Computer Science student at COMSATS University Islamabad. I work primarily at the intersection of AI systems and backend engineering — building things that need to handle real data, real users, and real edge cases.

I founded Zeploy Tech in 2026 to formalize the client work I was already doing. The company ships production software — e-commerce platforms, management systems, and custom tools — with a focus on clean architecture and reliable deployment.

My FYP, MockAI, is a multi-modal interview evaluation system that combines BERT-based NLP, DeepFace computer vision, and speech-to-text into a single analytics pipeline. It's the most technically complex system I've built so far.

I care about writing defensible APIs, designing database schemas that don't break under load, and shipping code that works in production — not just in demo environments.

EducationBS Computer Science — COMSATS University Islamabad, 2023–2027

LocationLahore, Pakistan

CompanyFounder & Developer — Zeploy Tech

FYPMockAI — AI Interview Evaluation System

SeekingBackend / AI Engineering roles

LanguagesEnglish (Fluent) · Urdu (Native)

CourseworkMachine Learning · Computer Vision · NLP · Database Systems

Technical Stack

Engineering Stack

Technologies used across production applications and AI-focused systems.

AI Systems

Machine Learning NLP Deep Learning BERT DeepFace OpenCV scikit-learn Pandas NumPy

Backend

FastAPI REST API Design JWT Auth API Integration

Frontend

React.js Flutter Tailwind CSS HTML5 CSS3 JavaScript

Databases

MongoDB Atlas MySQL Firebase

Tools

Git GitHub Postman Jupyter VS Code Linux Docker

Languages

Python JavaScript Dart Java C# SQL

Projects

Systems I've built

Production applications shipped for real clients and academic research.

01 / AI System In Development

MockAI

ChallengeEvaluate interview performance across speech, facial expression, and answer quality simultaneously — without manual review.

SolutionMulti-modal pipeline combining BERT NLP semantic grading, DeepFace emotion tracking, and speech-to-text transcription into a unified analytics dashboard.

React.js → FastAPI → BERT NLP → DeepFace → MongoDB

FastAPI React.js BERT NLP DeepFace MongoDB Atlas Python

Read Case Study → GitHub ↗

02 / E-Commerce Live

EduCart

ChallengeBuild a storefront with role-based admin controls and checkout flow that handles concurrent user sessions without data conflicts.

SolutionFull-stack platform with JWT session management, indexed SQL queries for catalog search, and transactional checkout logic with inventory locks.

React.js → FastAPI → JWT Auth → MySQL

React.js FastAPI MySQL JWT Auth

Read Case Study → Live ↗ GitHub

03 / Mobile App Internship Project

TuneBox

ChallengeHandle offline playback, audio buffer management, and seamless cloud/local source switching in a cross-platform mobile client.

SolutionReusable audio control library with Provider state management, local-first playlist caching, and async stream buffer handling. Reduced UI logic from ~1,200 to ~720 lines.

Flutter Dart REST APIs Provider

GitHub ↗

04 / Client System Client Deployment

Gym Management System

ChallengeReplace manual spreadsheet tracking for a gym facility — member enrollment, attendance logs, and subscription billing were error-prone.

SolutionOperations dashboard with transaction-safe enrollment workflows, automated attendance logging, and payment verification tracking.

FastAPI MySQL Python

Read Case Study → GitHub ↗

Experience

Where I've worked

Apr 2026 — Present

Zeploy Tech

Lahore, Pakistan

Founder & Full-Stack Developer

Architected, built, and shipped multiple production applications (including an e-commerce platform and gym management system) as a solo developer for real clients.
Led the entire product lifecycle from requirements gathering and system architecture design to database management, frontend creation, and cloud deployment.
Optimized database schemas and API endpoints, significantly reducing latency and eliminating redundant calls through query tuning and caching.

Sep 2025 — Apr 2026

Svype

Lahore, Pakistan

Software Development Intern

Developed major user interface flows and integrated RESTful API contracts for TuneBox, a Flutter-based music streaming app, using Provider state management.
Built a reusable audio control and playlist component library, optimizing performance and reducing frontend UI logic from ~1,200 to ~720 lines of code.
Partnered with cross-functional developers to design robust API contracts, resolve rendering inconsistencies, and run integration tests using Postman.

2024 — Present

Freelance

Remote

Backend & Full-Stack Developer

Designed and deployed custom web solutions for local clients integrating secure JWT authentication, FastAPI backends, and React.js interfaces.
Architected scalable RESTful APIs and optimized MySQL relational database schemas to handle high transaction volumes efficiently.
Implemented robust deployment pipelines and testing workflows to accelerate shipping frequency for client products.

At a Glance

By the Numbers

Production applications shipped

100+

Active users across deployed systems

AI subsystems in MockAI (NLP · Vision · Speech)

720

Lines of UI logic after refactor (down from 1,200)

Months to ship two client products solo

Industry certifications (Google AI · DeepLearning.AI)

Currently Building

What's Next

MockAI is in active development targeting Q4 2026 completion. Current focus: integrating a GenAI layer using the OpenAI API to generate personalised follow-up questions based on candidate responses — expanding the system from evaluation-only to a full AI interview coaching platform.

Client Reviews

Operational impact

"Syed Asjad built our operations panel on schedule and with rigorous database normalization. His focus on query optimization and transaction-safe checkout flows resolved membership synchronization errors we had struggled with for months."

Zia Ur Rehman Director of Operations, Gym Operations Panel

"Our biggest issue was double-allocation of stock during promotional sales. Syed integrated strict row locks at the database transaction boundary and cached our main catalog in Redis. Our endpoint latency dropped by 85%."

Abbas Haider Lead Product Owner, EduCart Platform

"Asjad contributed meaningfully to TuneBox during his time with us. He picked up our Flutter codebase quickly, delivered clean UI components, and his component library refactor noticeably reduced our frontend complexity. A self-sufficient developer who needs minimal hand-holding."

Wasif Ahmed CEO, Svype

Technical Writeups

Engineering Log

Detailed breakdowns of system bottlenecks, architectural choices, and performance optimizations from active production setups.

June 2026 · AI Systems

Building Multi-Modal Async AI Pipelines

Isolating deep-learning model evaluation threads using Python multiprocessing. Solving GIL-lock bottleneck and socket connection drops during real-time video/audio streaming under FastAPI.

Read Writeup →

May 2026 · Backend Engineering

FastAPI Query Tuning & Relational Indexing

Optimizing catalog queries on MySQL/MongoDB relation models. Implementing Redis caching layers and relational index tuning to reduce server execution latency under concurrent requests.