Sajjad Hussain
Sajjad Hussain

Most AI projects don't fail because of model choice — they fail because of dirty data and brittle pipelines. I build the boring infrastructure — proxy rotation, anti-detection scraping, ETL at scale — as carefully as the models sitting on top of it. That's the discipline I bring from data engineering into applied ML.

I'm a Software Engineering graduate (BS, CGPA 3.79/4.00) from The Islamia University of Bahawalpur. My final-year project combined Selenium/Scrapy scraping, ETL pipelines and Docker-deployed TensorFlow/PyTorch models behind a REST API — an AI-powered e-commerce recommendation system, end to end. That's still the template I work from: real data in, deployed system out.

Since 2022 I've freelanced on Fiverr while also serving as a Data Engineer at Code Apex, Islamabad. I'm currently exploring drift-aware cold-start recommendation systems as a research direction and looking to bring that same discipline to a full-time AI/ML or data engineering team or take it further at the graduate level.

Mission
Turn unstructured, real-world data into dependable, production-ready systems.
Vision
Grow into a full-time AI/ML engineering role at a team building frontier systems.
WORK PHILOSOPHY

How I approach a project

01
Understand the data first
Before writing a single line, I map the sources, the edge cases and what "clean" actually needs to mean for the use case.
02
Build for reliability
Retries, validation and de-duplication aren't afterthoughts they're what makes a pipeline usable in production.
03
Communicate constantly
Clear scoping and frequent updates are why clients come back — most of my Fiverr work is repeat business.