Early-career researcher · MS Artificial Intelligence
AI × Cybersecurity: building intelligent systems that detect threats, fraud and AI-generated content.
I am an early-career researcher with an MS in Artificial Intelligence. I build machine learning and NLP models that tell real from fake: AI-generated from human text, fraud from normal transactions, scams from genuine messages. Next, I want to make AI systems themselves secure.
whoami
Sana Ullah Khan · MS Artificial Intelligence
cat focus.txt
AI × Cybersecurity: phishing, scam & fraud detection, AI-generated text detection
ls research/
thesis/ (99.65% accuracy) · publications/ (3 peer-reviewed)
./open-to --list
Master's and PhD positions, research collaborations, and SOC / security and ML roles
Scored IELTS Academic 6.5 (B2) and started learning German.
Completed "IoT for Everyone" (UET Peshawar / HEC, Coursera).
01One focus: AI for security
Each area below is part of the same goal: using AI to detect and stop attacks, and keeping AI itself safe. Next to each one is what I can show for it today.
MS thesis on AI vs. human text classification with LSTM, BERT and hybrid models (99.65% accuracy, ~500,000 texts).
First author of a peer-reviewed paper on AI-driven fraud detection for FinTech cybersecurity.
Detecting AI-generated content today; adversarial ML and LLM security are what I want to study next.
Cisco Junior Cybersecurity Analyst career path and a 3-month NAVTTC course in digital forensics & cyber security (Grade A).
Cisco Ethical Hacker programme; hands-on lab work with Kali Linux, Nmap and Wireshark.
02About
I grew up in Chaman, Balochistan. My path has not been a straight line. After school I completed a three-year diploma in Health, Safety & Environmental Engineering, which taught me to think in terms of risk, hazards and prevention. I now apply the same way of thinking to digital systems.
{"I then studied Information Technology at BUITEMS in Quetta, where I took my first security and networking courses. \"My bachelor's grades don't show my best work"=>"I spent much of that time on cricket, football and freelance work. Towards the end of my BS I learned to manage my time and take my studies seriously.\""}
My MS in Artificial Intelligence at The Islamia University of Bahawalpur was the turning point. I graduated with a CGPA of 3.66, published three peer-reviewed papers with different teams, and wrote a thesis on detecting AI-generated text. At the same time I taught undergraduate IT courses as a visiting lecturer.
My university offered no cybersecurity track, so I built those skills myself through Cisco's Junior Cybersecurity Analyst and Ethical Hacker programmes and a NAVTTC course in digital forensics. The motivation is also personal: I have been the victim of an online scam and of a ransomware attack, and I see how weak cyber protection still is in Pakistan. AI is making attacks faster and more convincing, and I want to work on the side that defends against them.
03Research
MS thesis · 2026
I built and compared deep learning models that separate AI-generated from human-written text on a balanced dataset of about 500,000 samples. A paper based on the thesis is in preparation.
Models compared
Artificial Intelligence–Driven Fraud Detection in FinTech: Strengthening Cybersecurity Against Digital Financial Scams
Artificial Intelligence and Workplace Safety in the Oil and Gas Industry: Transforming Occupational Risk Assessment through Smart Safety Analytics
GenAI Adoption in Entrepreneurial Education and Digital Entrepreneurial Intention
Summaries, PDFs and BibTeX All citations on Google Scholar
Working on now
Want to study next
04Projects
A tool that flags scam and phishing messages written in English, Urdu and Roman Urdu, the languages people in Pakistan actually receive them in.
My MS thesis project. Deep learning models that separate AI-generated from human-written text, trained and compared on a balanced dataset of about 500,000 samples.
My BS final-year project, "IoT-based Pure Aqua Utilization in Smart Edifice". Arduino sensors and automatic pump control, with an Android app for tank level, water quality and billing.
Free tools
Small tools from my own security work. They run in your browser, with no sign-up.
How long would your password last against a real attack? Checks patterns, not just length.
Find out if a password has appeared in a data breach, without sending the password anywhere.
See where a link really goes and the warning signs of a phishing page, without opening it.
Spot fake domains that use look-alike letters (аpple.com), decode punycode, and list typosquats of your domain.
See the hidden GPS location, camera and date in a photo, and download a clean copy without them.
Find invisible characters, look-alike letters and hidden prompt injection in text before you paste it into AI.
05Experience
Mar 2024 – Jun 2026
Jul 2021 – present
06Skills
Pick an area to see what backs it up.
Research level
MS thesis on AI vs. human text classification with LSTM, BERT and hybrid models (99.65% accuracy, ~500,000 texts).
Lab and hands-on training level
07Teaching & service
Introduction to Information & Communication Technologies
Computer Networks
08Certificates
Digital Forensic & Cyber Security (3-month course, Grade A)
Junior Cybersecurity Analyst Career Path
Ethical Hacker
Introduction to Cybersecurity
Programming in Python
Advanced Chatbots with Deep Learning and Python
AI Python for Beginners
IoT for Everyone
IOSH Managing Safely
30-hour General Industry Safety and Health
09Education
Nov 2023 – Jan 2026
CGPA 3.66 / 4.00 · research thesis on AI vs. human text classification, supervised by Prof. Dr. Najia Saher.
Oct 2019 – Jul 2023
Courses in Information Security, Security Management, Cyber Security, Computer Networks and Systems & Network Administration. Final-year project on an IoT water-management system (Arduino sensors, automatic pump control, Android app).
2017 – 2020
85.08%, Grade B+. Taught me to think in terms of risk, hazards and prevention.
Contact
Open to Master's and PhD positions, research collaborations, and SOC / security and ML roles.