Hello there, I’m Abdullah, an aspiring Ph.D. candidate!
- abdullah [dot] ghani [at] lums [dot] edu [dot] pk
- abdullahghani2002 [at] gmail [dot] com
I am broadly interested in networked systems. Recently, I graduated from the Lahore University of Management Sciences, Pakistan with a 4.00/4.00, joint-highest in my batch (1/274), advised by Dr. Ihsan Ayyub Qazi and Dr. Zafar Ayyub Qazi.
Previously, I have worked extensively on building a framework to longitudinally investigate how online trackers on sensitive websites compromise user privacy (Accepted @ IMC 2026, advised by Dr. Zubair Shafiq), conducting an advertiser-level audit across 30 countries to evaluate global disparities in child-safety practices in children’s video advertising and building a scalable method to mitigate associated risks (under review, AAAI ICWSM), and examining the risks and opportunities of large language models (LLMs) in multilingual, global fact-checking (Accepted @ Nature Scientific Reports).
I am currently seeking Ph.D. positions for Fall 2026.
Publications
PixelConfig: Longitudinal Measurement and Reverse-Engineering of Meta Pixel Configurations
Abdullah Ghani, Yash Vekaria, Zubair Shafiq.
ACM Internet Measurement Conference (IMC) 2026.
PDF | Code | Tool
Scaling Truth: The Confidence Paradox in AI Fact-Checking
Ihsan Ayyub Qazi, Zohaib Khan, Abdullah Ghani, Agha Ali Raza, Zafar Ayyub Qazi, Wassay Sajjad, Ayesha Ali, Asher Javaid, Muhammad Abdullah Sohail, Abdul Hameed Azeemi
Nature Scientific Reports.
PDF | Code | Data
Advertisers, Provenance, and Safety: A 30-Country Audit of Children’s YouTube Ads
Abdullah Ghani, Yahya Khawaja, Usman Naseem, Ihsan Ayyub Qazi, Zafar Ayyub Qazi.
Under Review.
Other Research Projects
In addition to the aforementioned projects, I have also contributed to the following projects:
Embedding The Truth: Approximate Caching for Fact Checking
Course Project for CS6303: Topics in Large Language Models. Explored the feasibility of reusing fact-checking claims across organizations to improve efficiency. Used an approximate caching approach with multilingual embeddings in a vector database to identify recurring claims. [PDF]Decoding User Concerns in AI Health Chatbots
Course Project for CS682: Topics in Computer and Network Security. Collected and analyzed over 50,000+ user reviews from health chatbot apps using BART and Gemini to assess their effectiveness compared to manual analysis and to qualitatively examine user concerns related to data misuse, transparency, and consent. [arXiv Link] | [Got Cited by Google!]
Current & Ongoing Research
As the lead of the Internet, Data, and Society lab, I am currently working on the following problems:
- Fidelity of LLMs in Clinical Settings: Evaluating the clinical fidelity of Large Language Models on real-world medical vignettes.
- Disinformation Dynamics in Geopolitical Conflicts: Conducting a systematic, cross-platform analysis of the spread and engagement dynamics of disinformation during a geopolitical conflict.
- Reasoning and Bias in LLMs as Participants: Investigating the reasoning behaviors and regional biases manifested when using LLMs as participants.
Development & Software Projects
Beyond my formal research, I enjoy building software that solves interesting technical challenges and deepens my understanding of core computer science principles.
Distributed, Fault-Tolerant Key-Value Store
(Go): Engineered a distributed data store from scratch, implementing the Raft consensus algorithm to achieve fault tolerance through leader election and persistent log replication. [GitHub]Virtual UNIX File System
(C): Implemented a UNIX-like virtual file system with a partitioned architecture of superblocks, inodes, and data blocks to manage low-level file I/O operations. [GitHub]User-Level Threading Library
(C): Developed a lightweight threading library that enables application-level context switching by directly managing CPU registers and process control blocks (PCBs). [GitHub]Succession Planning Software
(MERN Stack, TensorFlow): Designed and built a full-stack application that uses a TensorFlow model to analyze performance data and identify high-potential candidates for leadership roles. [GitHub]PakStay Advisor
(Python, Scikit-learn): Created an end-to-end data science project that scrapes hotel data and uses predictive models to provide profitability and reliability insights for both travelers and hotel managers. [Blog]
Teaching Experience
- Topics in Large Language Models: Systems and Applications (Ongoing)
- Data Science (Fall 2024)
- Data Structures (Spring 2024)
- Discrete Mathematics (Fall 2023)
Apart from solidifying my own understanding of these subjects in unanticipated ways, taking on the role of a teaching assistant has allowed me to:
- Creatively design programming assignments (e.g., AVL Trees in C++, Causal Inference in Python).
- Structure comprehensive exams and quizzes.
- Deliver live tutorials and supervise semester-long data science projects.
- Use code-similarity checkers (e.g., JPlag) to maintain academic integrity.
Awards and Honours
- Full Merit Scholarship Award: Granted to the top 1% students from LUMS SBASSE Batch of 2025 from the 2022-2025 academic years.
- Dean’s Honour List: Consecutively awarded from 2021-2025 academic years.
- Top Candidate at School (A Levels): Achieved the highest grades at Lahore Grammar School Defence.
- Top 20 National Mathematics Talent Contest: Shortlisted from candidates across Pakistan to represent Pakistan at the International Mathematics Olympiad.
- Academic Excellence: Scored an A+ in over 70% of my graded courses.
Miscellaneous
- A very helpful resource for graduate school applications that greatly aided me in my journey.
- Some golden advice from my mentor, Dr. Ihsan, that I strive to follow in my academic and personal growth.
- I wish I could find the time to swim everyday, it does wonders.
- When I’m not working, I love exploring and spending time alone with nature. Here is a photo from a recent, arduous trek I made to Chitta Katha Lake in the Himalayas!
Enroute to Chitta Katha Lake, Shounter Valley.