Profile banner
Neil Natarajan

Neil Natarajan

Human-centered AI researcher (DPhil, University of Oxford)

neil_enna
Pittsburgh, Pennsylvania
Joined February 2025
Favicon imageneilenatarajan.com
Favicon imageneilenatarajan.com
Favicon imagecs.ox.ac.uk
Favicon imagescholar.google.com
Favicon imageuk.linkedin.com

Network

1.1K connections
⚖️
AI Research Policy
🤖
AI Founders Leaders
❤️
Impact Philanthropy
💰
Startup Investors
🏛️
Ellison Institute Oxford
📊
Data Science Analytics
🧬
OHSU Research

Summary

Human‑centred AI researcher focused on the intersection of HCI, explainable AI, and selection systems. His doctoral work and consulting projects revolve around designing AI decision‑support that is informed by participatory design, fairness thinking, and empirical evaluation in real selection contexts (e.g., scholarship selection and Schmidt Futures' Rise program). neilenatarajan+2
Applied evaluation of generative‑AI detectors and practical tradeoffs: he has led case‑study evaluations of tools like GPTZero, documenting calibration, accuracy, and subgroup bias, and arguing detectors are useful for aggregate insight but not for high‑stakes disqualification without careful calibration. github+1
Bridges research and engineering: alongside HCI and fairness research, he has hands‑on experience in ML/data‑science internships and web/frontend engineering (BigID, Babbel), and contributed to building and auditing an automated research system at Autoscience. neilenatarajan+1
Teaching and mentoring experience at Oxford: served as a Stipendiary Lecturer at Balliol College, delivering tutorials and undergraduate instruction in computer science, indicating strength in explaining technical concepts and supervising students. neilenatarajan+1

Work

Education

Projects

Writing

Detecting Generative AI Usage in Application Essays

January 1, 2024

Case-study evaluation of GPTZero and other detection approaches in real scholarship selection settings; finds detectors useful for aggregate insights but insufficiently reliable for high-stakes disqualification due to sensitivity and heterogeneous bias.

Favicon imagegenerativeaiandhci.github.io

Human-AI Collaboration in Recruitment and Selection (Doctoral Consortium, IJCAI 2023)

January 1, 2023

Doctoral-consortium piece exploring how AI decision-support tools and human selectors interact in recruitment and selection contexts.

Favicon imageneilenatarajan.com

Trust explanations to do what they say

January 1, 2022

Workshop / preprint work on the limits and appropriate uses of explanations for AI decisions (human-centered perspective on XAI).

Favicon imagearxiv.org

Variation in an iron metabolism gene moderates the association between blood lead levels and ADHD in children

January 1, 2016

Co-authored empirical paper in Psychological Science (2016) on gene–environment interactions and ADHD outcomes; demonstrates earlier interdisciplinary research contributions.

Favicon imagescholar.google.com

Hobbies

Reading, rock climbing, badminton, and pickleball. neilenatarajan