AI / ML Engineer & Researcher

Taran Rai

AI/ML engineer working across computer vision, deep learning, LLMs, and applied AI. I build, research, and ship things that work in the real world.

Taran Rai
Scroll

What I do

I do AI/ML engineering and data science things. My work spans computer vision, deep learning, LLMs, and applied AI/ML. I build and validate algorithms for real-world deployment, and I stay close to the cutting edge of what's happening in AI.

A lot of my research has been in computational pathology and the health domains (both animal and human), but I've also worked across social media listening with pharma companies, wearable tech and sensor-based systems through the vHive Animal Health Incubator, and I'm interested in the human-technology interaction side of AI. My experience spans niche startups through to larger multinational organisations.

I guest lecture for AI and AI Programming at the University of Surrey, introducing neural networks.

I hold a PhD in AI/ML for digital pathology (deep learning + computer vision for grading in canine sarcoma biopsy images), an MSc in Data Science from City, University of London, and a BA (Dual Hons) in Economics and Finance from Keele University.

Some of my passions

My interests are shaped by personal experiences. I choose to work on projects that genuinely matter to me.

🔬

Computational Pathology

Applying and developing AI/ML models for animal and human health. Most of my research lives here.

🌾

Farming & Nature

Protecting and providing for future generations. Farming, planting lots of trees, and nurturing the land.

💸

Investment Accessibility

Collaborating with investment managers from boutique investment banks to make investing easier to understand and more accessible. Watch this space.

The LiverSWAT Project

After losing my father to liver cancer in 2024, we created LiverSWAT at NishkamSWAT, fundraising for FibroScan liver screening roadshows across the UK.

Selected Publications

Full list available on Google Scholar.

2026
Journal of Imaging Informatics in Medicine, 1-24
Mirhosseini S, Rai T, Diaz-Santana P, La Ragione R, Bacon N, Wells K
2026
Cureus Journal of Computer Science, 3(1)
Alkhadra RB, Diaz-Santana P, Rai T, Wells K
2025
SPIE Medical Imaging 2025
Mirhosseini S, Rai T, Diaz-Santana P, La Ragione R, Bacon N, Wells K
2024
SPIE Medical Imaging 2024
Rai T, Gola C, Hernández M, Fingerhood S, Marrero J, Diaz-Santana P, Giglia G, Morisi A, Bacci B, Thomas SA, Ressel L, Bacon N, Papachristou N, Cook A, La Ragione R, Wells K
2024
Cancers, 16(3), 644
Rai T, Morisi A, Bacci B, Bacon NJ, Dark MJ, Aboellail T, Thomas SA, La Ragione RM, Wells K
2023
Deep Learning for Necrosis and Mitosis Detection in Canine Soft Tissue Sarcoma Whole Slide Images
PhD Thesis, University of Surrey
Rai T
2023
Seminars in Oncology Nursing
Papachristou N, Kotronoulos G, Dikaios N, Allison SJ, Eleftherochorinou H, Rai T, Kunz H, Barnaghi P, Miaskowski C, Bamidis PD
2023
Veterinary Sciences, 10(1), 45
Morisi A, Rai T, Bacon NJ, Thomas SA, Bober M, Wells K, Dark MJ, Aboellail T, Bacci B, La Ragione RM
2023
Investigating the Potential of Untrained Convolutional Layers and Pruning in Computational Pathology
Medical Imaging 2023: Digital and Computational Pathology, SPIE, Vol. 12471
Rai T, Papanikolaou I, Dave N, Morisi A, Bacci B, Thomas S, La Ragione R, Wells K
2022
Scientific Reports, 12, 10634
Rai T, Morisi A, Bacci B, Bacon NJ, Dark MJ, Aboellail T, Thomas SA, Bober M, La Ragione R, Wells K
2019
Can ImageNet Feature Maps Be Applied to Small Histopathological Datasets for the Classification of Breast Cancer Metastatic Tissue in Whole Slide Images?
Proc. SPIE, 10956, 109560V
Rai T, Morisi A, Bacci B, Bacon NJ, Thomas S, La Ragione R, Bober M, Wells K

Talks & Appearances

18 Mar 2025
25 Feb 2025
City St George's Alumni Network
Madicken Malm
21 Jul 2023
Vet Record
Georgina Mills
05 May 2023
07 Dec 2022
09 Sep 2019

Model Takes

My takes on AI/ML and technology. Sometimes other things I find interesting.

Loading posts...

Get in touch

Open to interesting work, collaborations, and conversations.