About
I’m a Senior Researcher in the Modernising Medical Microbiology (MMM) group, Nuffield Department of Medicine, University of Oxford, and an Honorary Researcher at Oxford University Hospitals NHS Foundation Trust.
I work on developing and evaluating safe, reliable, and practical machine learning systems for healthcare, with a focus on infectious diseases, antimicrobial resistance, and public health surveillance. Current projects include extracting information from free-text clinical reports and building on-device language models for privacy-preserving clinical applications. More broadly, I’m interested in clinical and small language models, model evaluation, EHR modelling, and GFMs.
Before Oxford, I spent nearly six years at M42 in Abu Dhabi, working on foundation models for healthcare and biology (Med42, BioFM, CXformer) and on evaluating clinical LLMs (MEDIC). Outside work, I enjoy hiking, travelling, and Brazilian Jiu-Jitsu.
News
- MEDIC, our framework for evaluating clinical LLMs, is published in TMLR.
- BioToken & BioFM accepted at ICML 2026.
- Cross-Examination Framework accepted at ACL 2026.
- Joined the University of Oxford as Senior Researcher in healthcare AI.
Show older news
- Tokenization to Transfer accepted at ICLR 2026.
- Building Trust in Clinical LLMs accepted at EMNLP 2025.
- Talk on MEDIC at NVIDIA GTC 2025.
Experience
Selected Publications
Full list on Google Scholar. * equal contribution.
MEDIC: Comprehensive Evaluation of Leading Indicators for LLM Safety and Utility in Clinical Applications
P. Kanithi, C. Christophe, M. A. Pimentel, T. Raha, P. Munjal, N. Saadi, H. A. Javed, … S. Khan
TMLR 2026 · PDF · Leaderboard
BioToken and BioFM: Biologically-Informed Tokenization Enables Accurate and Efficient Genomic Foundation Models
A. Medvedev*, K. Viswanathan*, P. Kanithi*, K. Vishniakov, P. Munjal, C. Christophe, … S. Khan
Cross-Examination Framework: A Task-Agnostic Diagnostic for Information Fidelity in Text-to-Text Generation
T. Raha, C. Christophe, N. Saadi, H. A. Javed, M. A. Pimentel, R. Rajan, P. Kanithi
Tokenization to Transfer: Do Genomic Foundation Models Learn Good Representations?
K. Vishniakov, K. Viswanathan, A. Medvedev, P. Kanithi, M. A. Pimentel, R. Rajan, S. Khan
Building Trust in Clinical LLMs: Bias Analysis and Dataset Transparency
S. Maslenkova, C. Christophe, M. A. Pimentel, … P. Kanithi
Empirical Analysis of Scaling Vision Foundation Models for Chest X-rays
A. Al-Mahrooqi, P. Munjal, R. Rajan, M. A. Pimentel, P. Kanithi
Beyond Fine-tuning: Unleashing the Potential of Continuous Pretraining for Clinical LLMs
C. Christophe, T. Raha, S. Maslenkova, M. U. Salman, P. Kanithi, M. Pimentel, S. Khan
Findings of EMNLP 2024 · PDF
Education
Interactive image segmentation of MARS spectral CT datasets
Immersive augmented reality system for assisting needle positioning during ultrasound-guided intervention