Jiri Gesi bio photo

Jiri Gesi

Machine Learning | Retrival & Ranking researcher

Twitter   G. Scholar LinkedIn Github e-Mail

We have openings for Applied Scientist Interns and full-time positions! Contact me if you’re interested in working on LLM for retrival&ranking!

About me

I’m currently an Applied Scientist at Amazon Search (formerly A9), where I specialize in developing Large Foundation Models. My work involves using advanced AI techniques to pre-train and fine-tune Large Language Models (LLMs) to understand customer behaviors and product characteristics. This helps us unlock the full potential of LLMs for creating better retrieval and ranking models for both organic search results and advertisements. I hold a PhD in Representation Learning of Programming Languages from UC, Irvine.


I love traveling, hitting up music festivals, and saying ‘Yes!’. I enjoy coding and reading research papers.

My research

The embedding or dense vector representation of text captures its semantic information, enabling its use in various downstream applications such as retrieval, reranking, classification, clustering, and semantic textual similarity tasks. This method is essential for embedding-based retrieval systems, which play a crucial role in retrieval-augmented generation (RAG). RAG allows large language models (LLMs) to access the latest external or proprietary knowledge without altering the model parameters. And my research focuses on:


How can we train AI systems to efficiently and accurately represent the human behaviors and products for LLM?


My team at Amazon Science is researching how to train large Natural Language Processing models to generate embeddings that represent customer online behaviors for organic and advertisement ranking. We’re also developing embeddings for store products to enhance the accuracy and efficiency of our ‘ChatGPT for shopping’ agent, among other projects.

Professional Services

Publication reviewer:

  • (NIPS 2024) The Thirty-eighth Annual Conference on Neural Information Processing Systems
  • (TIST 2024) Transactions on Intelligent Systems and Technology 2024
  • (WWW 2024) The 2024 International World Wide Web Conference
  • (AMLC 2024) The 2024 Amazon Machine Learning Conference
  • (ASE 2024) The 39th IEEE/ACM International Conference on Automated Software Engineering
  • (TOSEM 2024) Transactions on Software Engineering and Methodology 2024
  • (MSR 2024) 21st International conference on Mining Software Repositories
  • (SEA4DQ 2024) The 4th International Workshop on Software Engineering and AI for Data Quality in Cyber-Physical Systems/Internet of Things
  • (TOSEM 2023) Transactions on Software Engineering and Methodology 2023
  • (ESE 2021) Empirical Software Engineering Journal, 2021

Program co-chair/committee member:

  • (PC member) The 39th IEEE/ACM International Conference on Automated Software Engineering (ASE 2024)
  • (PC member) The 21st International conference on Mining Software Repositories (MSR 2024)
  • (Industry Chair) The International Workshop on Software Engineering and AI for Data Quality in Cyber-Physical Systems/Internet of Things (SEA4DQ 2024)
  • (PC member) ACM International Conference on AI-powered Software (AIware 2024)

Personal News

  • 03/2024 📢 Our paper Towards Robustness Analysis of E-commerce Ranking System is accepted in the WWW (The Web) 2024 Conference.
  • 09/2023 📢 Jiri joined Amazon A9 (Search Science and AI) as an Applied Scientist
  • 12/2022 📢 Our paper “Leveraging Feature Bias for Scalable Misprediction Explanation of Machine Learning Models” is accepted in ICSE 2023
  • 06/2022 📢 Jiri interned at Amazon A9 (Search Science and AI) as an Applied Scientist Intern
  • 01/2022 📢 Jiri got PhD research fellowship funding from eBay for 2022
  • 12/2021 📢 Jiri advances to Ph.D. candidate at the school of Information and Computer Science, UC Irvine
  • 04/2021 📢 Our paper “An Empirical Examination of the Impact of Bias on Just-in-time Defect Prediction” is accepted in ESEM 2021
  • 03/2021 📢 Our paper “Evaluating and Improving Static Analysis Tools Via Differential Mutation Analysis” is accepted in QRS 2021
  • 01/2021 📢 Jiri got PhD research fellowship funding from eBay for 2021
  • 09/2019 📢 Jiri is awarded 2019 Chair’s Award from Donald Bren School of Information and Computer Sciences Department

Awards

  • 🏆 Dean’s Awards from Donald Bren School of Information and Computer Sciences,UC Irvine, 2019
  • 🏆 Outstanding Master student’s thesis, University of Michigan, 2017
  • 🏆 Siyuan Scholarship, Xi’an Jiao Tong University, 2013, 2014, 2015

my FULL CV.


This site has been vistited Hits daily/total times since Feb 1, 2022.