Sagar Uprety

Senior Research Data Scientist, University College London (UCL), UK

 Professional biography

I am a Senior Research Data Scientist in the Advanced Research Computing Centre (ARC) in University College London (UCL), UK.  I work primarily in research and development of AI models and systems, particularly pertaining to the Natural Language Processing (NLP) domain. I have experience in Document AI - information extraction from visual documents, and Large Language Model (LLM) applications.

I am also a Visiting Fellow at School of Computing and Communications, The Open University, UK where I collaborate with Prof. Dawei Song on quantum-inspired Information Retrieval models and work in the OpenMars project to build time-series algorithms for weather forecasting on Mars. 

I have a background in Information Retrieval (IR) and NLP,  User Behaviour Modelling , Time-series forecasting and Deep Reinforcement Learning (DRL).  Previously, I worked as a Research Scientist at Bravura Solutions,  a leading provider of financial software solutions, with a focus on Document AI and LLMs.

 Before Bravura, I worked as a Postdoctoral fellow at  Warwick Manufacturing Group, University of Warwick, UK, focusing on DRL algorithms for complex robotic manipulation tasks and multi-agent learning . 

 From 2017-2020, I held the Marie Skłodowska Curie Early Stage fellowship to pursue PhD at The Open University, UK finishing in August 2020.  My doctoral research, as part of the  European Commision's Horizon 2020 QUARTZ project,  focused on modelling user behaviour in search and information interaction using the quantum mathematical framework.  Within this project, I was also been seconded twice to Queensland University of Technology, Australia and worked with Prof. Peter Bruza's Information Systems group

I have a Bachelor's degree in Computer Science and a Master’s in Physics from BITS Pilani, India. After graduating, I worked for 2 years as a Software Engineer with projects including writing backend code for a vertical search engine (Stayzilla) and to perform query intent classification  and information extraction for voice search and chatbots using Deep Neural Networks (at Justdial). 

Research interests

Interesed in solving real-world problems through machine learning. Particularly interested in building products based on large language models (LLMs).

Publications

A list of my publications can seen at Google scholar. 

Blog

[1]   Data Cleaning and Exploratory Analysis in Python and R   

[2] Langchain Agents: Prompt Design for Tools Control Flow



Contact Details:

Email: sagaruprety@gmail.com

LinkedIn: https://www.linkedin.com/in/upretysagar/