cv

Download CV

Research Statement

I aim to build open source frameworks that create real world impact while advancing academic research. I focus on transforming digital advertising through automated content moderation and detecting malicious actors in the ad supply chain. My past work includes empirical studies on privacy-enhancing technologies, propaganda, and harmful ads; using AI to detect contextual nuances and mitigate these issues at scale. Going forward, I aim to enhance user experience online by uncovering differential treatment, fighting fraud, and conducting large scale measurements to improve widely used tools.

Education

  • Ph.D. in Computer Science (Sep 2021 - Jun 2026)
    • New York University (NYU), New York City, USA
    • Advisors: Prof. R. Greenstadt & Prof. B.D. Gavitt
    • Thesis: Adblocking’s Privacy Tradeoffs and the Future of Sustainable Advertising
  • B.Tech in Computer Science and Engineering (Sep 2017 - May 2021)
    • Indian Institute of Technology Bombay (IITB), Mumbai, India
    • GPA: 8.15/10.0
    • Thesis: Large-scale assessment of vulnerabilities in open-source network server binaries

Work Experience

  • Sr. Staff Web Security Researcher, Palo Alto Networks (Aug 2026 - Present)
    • Santa Clara, US Guide: Oleksii Alex Starov
    • Part of the Advanced URL Filtering team handling malicious web URLs including JS, HTML, etc.
  • PhD Software Intern, Uber Technologies (Sep - Nov 2025)
    • Sunnyvale, US Guides: Xandra Xhu & Bo Ling
    • Implemented an LLM-based generative recommender model for UberEats homefeed for tar-aware predictions
    • Developed TPU and GPU compatible frameworks for generative recommender models for efficient hardware benchmarking, achieving 5.3x higher throughput on TPUs
  • Research Intern, CISPA Helmholtz Center for Information Security (Jun - Aug 2024)
    • Saarbruecken, Germany Guide: Ben Stock
    • Developed a novel mechanism to identify the differential treatment of adblocker users by websites, uncovering potential for fingerprinting users and degrading their user experience
    • Instrumented Google Chrome’s V8 engine to collect JS execution logs to understand its inner workings
  • Research Intern, University of California, Santa Barbara (Apr - Nov 2020)
    • Santa Barbara, USA Guides: Giovanni Vigna and Christopher Kruegel
    • Developed KANF, a kernel-assisted network fuzzer, using Linux kernel driver modules and networking tools to test over 10,000 open-source network programs and conduct bug detection at scale
    • Interleaved the Linux Kernel with AFL using kernel driver modules and network programs
  • Software Engineer Intern, A.P.T Portfolio (Apr - Jun 2020)
    • Delhi, India Guide: Pratyush Rathore
    • Reported and patched crucial bugs in the source code implemented for processing daily traffic in excess of 4 crore orders at NSE, and developed and optimized a dynamic latency based exchange simulation model
  • Cyber Security Research Intern, Lucideus (May - Jul 2019)
    • Delhi, India Guide: Rahul Tyagi
    • Hardened CentOS Linux using 239 remediations as provided by CIS (Center for Internet Security) and prepared detailed documentation covering attacks and mitigation techniques on the OWASP Top 10

Selected Publications

  • CCS ‘26: AdLens: Efficient Detection of Deceptive Software Ads (w/ M.A. Darwish, M.A. Aghdam, R. Greenstadt, G. Acar) [Paper] [Code] [Website]
    • Designed an open-source, multilingual two-stage LLM pipeline that audits ad transparency libraries at scale, cutting inference cost 17x versus exhaustive classification while achieving 0.92 F1
    • Used it to expose thousands of deceptive software ads and the malvertising campaigns hidden behind them, leading to takedowns after disclosure to Google
  • PETS ‘25: Automated detection and evaluation of problematic ‘allowed’ advertisements (w/ Julia Jose, Hussam Habib, Rachel Greenstadt) [Paper] [Poster] [Code]
  • AIWILD, ICLR ‘26: When Agents Persuade: Rhetoric Generation and Mitigation in LLMs (w/ Julia Jose, Rachel Greenstadt) [Poster]
  • AsiaCCS ‘24: A User-Focused Evaluation of Privacy-Preserving Browser Extensions (w/ Rachel Greenstadt) [Paper] [Code]
  • SecWeb, IEEE S&P ‘24: Analysis of web breakages caused by adblockers (w/ Mitchell Zhou, Ben Stock, Rachel Greenstadt) [Paper] [Code]
  • USENIX ‘22: Drifuzz: Harvesting Bugs in Device Drivers from Golden Seeds (w/ Zekun Shen, Brendan Dolan-Gavitt) [Paper] [Code]
    • Implemented a framework for concolic fuzzing PCI device drivers, discovering and patching 12 bugs and obtaining 2 CVEs in the Linux driver code

Position of Responsibility

  • Teaching Assistant, New York University Application Security (Jan 2022 - Apr 2022)
    • Mentored a class of over 100 students in a remote setup, facilitating practical class demonstrations to instill a better understanding of niche concepts
  • Thesis Mentor, New York University — Real-time measurement of web breakages, Mitchell Zhou (Sep 2023 - May 2024)
    • Mentored an undergraduate student for his bachelor’s thesis on understanding web breakages and measuring them at scale, leading to a publication at SecWeb, IEEE S&P

Reviewer Duties

  • Program Committee: NDSS ‘25, PETS ‘24, MADWEB ‘25
  • Artifact Committee: CCS ‘24/25, USENIX ‘23/24, PETS ‘25

Awards / Leadership

  • Presented research posters at Columbia Privacy Day and Google Ad Privacy Day [2025]
  • Received a scholarship to attend summer school at EPFL, Switzerland and CISPA, Germany [2024]
  • Received SoE fellowship from NYU in freshman year to facilitate research goals [2021]
  • Secured All India Rank 48 in JEE-Advanced out of 220,000 shortlisted candidates [2017]
  • Awarded KVPY Fellowship and NTSE Scholarship by the Government of India [2016]

Skills

  • ML Paradigms: Natural Language Processing, Prompt Engineering, Sentiment Analysis, Thematic Analysis, Supervised/Unsupervised Learning, RAG
  • ML Frameworks: PyTorch, LLMs, VLMs, NumPy, Pandas, Jupyter
  • Internet Measurement: Puppeteer, Selenium, Playwright, Web Extensions
  • Languages: C/C++, Python, Bash, Java, Assembly, JavaScript
  • Security Tools: Kali Linux, Metasploit Framework, Xerosploit, Reversing Tools
  • Software Tools: Linux, Git, MATLAB, MySQL, AutoCAD, LaTeX, AWS, HPC, Slurm