Shreshth Tuli
Director Of AI (Model Orchestration)
London, England, United Kingdom
Summary
👤
Senior🎓
Top SchoolShreshth Tuli is a Director of AI specializing in model orchestration who builds hybrid, agentic architectures that span on-device, edge and cloud environments to deliver adaptive intelligence where it matters. With a PhD in AI and optimization from Imperial College London and eight years of experience across academia, industry and government collaborations, he blends deep research in symbolic learning and optimization theory with hands-on systems work—ranging from anomaly-detection transformers (contributions to the TranAD pipeline) to early-exit LLMs for mobile. He has applied AI to practical domains including autonomous robotics, drug discovery, and customer lifecycle optimization, and plays active roles shaping 6G+AI standards and AI special interest groups. Recognized among Stanford’s top 2% of scientists, he pairs rigorous publication track-record with product-focused delivery at organizations from IIT Delhi and Meta to Lenovo. Less obvious: he repeatedly moves between research and production roles, making him adept at translating novel algorithms into robust, resource-aware deployments across heterogeneous compute.
8 years of coding experience
7 years of employment as a software developer
Indian Institute of Technology Delhi (IIT Delhi)Doctor of Philosophy - PhD Artificial Intelligence and Optimization Theory, Doctor of Philosophy - PhD Artificial Intelligence and Optimization Theory at Imperial College LondonLondon School of Economics and Political ScienceCBSE Science, CBSE Science at Amity International School, Saket
English, Hindi, French
Github Skills (7)
data-preprocessing10
pandas10
machine-learning10
anomaly-detection10
python10
numpy10
unsupervised-learning9
Programming languages (5)
Github contributions (5)
imperial-qore/TranAD
Mar 2021 - Jan 2023
[VLDB'22] Anomaly Detection using Transformers, self-conditioning and adversarial training.
Role in this project:
Data Scientist
Data ScientistContributions:115 commits, 1 PR, 21 pushes in 1 year 11 months
Contributions summary:Shreshth contributed to the data preprocessing stage by implementing data loading and saving functions for different datasets, including synthetic and SMD datasets. The user added normalization steps to the data processing pipeline. Further contributions involved updating the file names in the processing pipelines and adding the creation of a dataset for SMAP and MSL.
adversarial-learningpytorchanomalytransformer-modelstransformers
imperial-qore/MetaNet
Nov 2021 - May 2022
Contributions:131 commits, 1 push in 5 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.