Tanay Aggarwal, a PhD student at KMi, has joined Amazon as an Applied Scientist II for a 16-week internship in Seattle, where he will apply his doctoral research on artificial intelligence and information retrieval in an industrial setting.
Tanay’s PhD, supervised by Dr Angelo Salatino, Dr Francesco Osborne, and Prof Enrico Motta, investigates how large language models can perform structured reasoning over research disciplines, enabling cross-disciplinary knowledge modelling, semantic link prediction, and scalable knowledge graph construction. His broader aim is to advance trustworthy AI systems that improve knowledge discovery, representation, and personalised recommendations for research and education.
Tanay has developed new benchmarks for semantic relationship inference and studied how reliably both open-source and proprietary models perform across various research disciplines. His research has been published in venues including Information Processing & Management, ACM Transactions on Intelligent Systems and Technology and Quantitative Science Studies, and presented at the International Semantic Web Conference and the European Semantic Web Conference.
(Image caption: Tanay Aggarwal at The Spheres, Amazon’s headquarters campus in Seattle, WA.)
Bringing research rigour into industry
The questions at the heart of Tanay’s PhD, how to retrieve relevant knowledge, how to make LLM reasoning reliable, and how to measure whether results can be trusted, are equally central to applied science in industry, where they must be answered at far greater scale.
Tanay was selected following interviews with senior applied scientists. “What struck me during the interviews was how much they valued rigour,” said Tanay. “Being able to explain where my models fail, not just where they succeed, mattered as much as the results themselves. That habit of honest evaluation is something I owe directly to KMi and my supervisors.”
Strengthening links between research and practice
During the internship, Tanay hopes to gain first-hand experience of how research methods become dependable production systems, from evaluation at scale to the day-to-day collaboration between scientists and engineers. He will return to KMi afterwards to complete his doctorate, bringing those insights back into the final stages of his thesis.
“The gap between academic research and industrial practice is often narrower than it looks,” Tanay added. “I’m grateful to my supervisors, and to Prof John Domingue and Dr Tracie Farrell for their mentorship.”
Placements like this highlight the real-world relevance of KMi’s research and help build lasting connections between academia and industry.