# Sarath Shekkizhar, Ph.D. Staff Research Scientist at Salesforce. ## Education ### Ph.D. in Electrical Engineering University of Southern California, Aug 2017 - May 2023. GPA: 3.93. Advisor: Antonio Ortega. ### M.S. in Computer Science University of Southern California, Aug 2017 - May 2022. GPA: 4.0. ### M.S. in Electrical Engineering (Computer Vision, Machine Learning) University of Southern California, Aug 2012 - Dec 2013. GPA: 3.86. ### B.Tech. in Electronics and Communication National Institute of Technology, Tiruchirappalli, July 2008 - June 2012. GPA: 9.12. ## Work experience ### Staff Research Scientist, Salesforce Oct 2024 - Present, San Francisco, CA. Working on foundational research on (multi) agentic systems and LLM training for improved reasoning and alignment. I continue to spend some of my research time on aspects of voice AI and agentic design for applied research. ### Member of Technical Staff, Tenyx (Acq. by Salesforce) June 2023 - October 2024, Los Altos, CA. Part of the founding team at Tenyx building Voice AI for customer support. Was primarily focused on research and algorithms for various aspects of voice agents. Key accomplishments include research on continual learning, building TenyxChat series of models, and geometric characterization of LLMs. Was also involved in product development, particularly in endpointing, audio disambiguation, and agent governance. ### Research Intern, Google Sep 2022 - Dec 2022, Sunnyvale, CA. Worked on understanding the impact of input data used in training graph models and scalable sampling approaches to improve semi-supervised graph learning. Preliminary experiments with proposed graph learning showed 3x increased recall in abuse detection. Host: Mohamed Farghal, Animesh Nandi, Behavior Protections, Counter-Abuse Technology. ### Software Engineer 2, KLA Tencor Mar 2014 - Oct 2016, Milpitas, CA. Designed and developed tools to classify and visualize defect modulations for Process Window Qualification in wafer fabrication. Also, implemented and co-owned components for analysis and classification using decision trees and random forests. ## Selected publications - [Counterparty Modeling is Not Strategy: The Limits of LLM Negotiators](https://arxiv.org/abs/2605.16575) — arXiv Preprints, 2026 - [Beyond the Assistant Turn: User Turn Generation as a Probe of Interaction Awareness in Language Models](https://arxiv.org/abs/2604.02315) — arXiv Preprints, 2026 - [Interaction Theater: A case of LLM Agents Interacting at Scale](https://arxiv.org/abs/2602.20059) — arXiv Preprints, 2026 - [Echoing: Identity Failures when LLM Agents Talk to Each Other](https://arxiv.org/abs/2511.09710) — arXiv Preprints, 2025 - [Convergence dynamics of Agent-to-Agent Interactions with Misaligned objectives](https://arxiv.org/abs/2511.08710) — arXiv Preprints, 2025 See the [complete publications index](https://shekkizh.com/publications). ## Patents - Knowledge base for voice large language model applications — US63752613, Provisional - Gradient-free optimization of large language models — US63752618, Provisional - Machine learning model compression — US18905761, Provisional - Training a target activation sparsity in a neural network — US18802235, Pending - Domain aware large language model governance — US18745562, Granted - Fine-tuning machine learning models while retraining accumulated knowledge — US18496698, Pending - Data sampling using Locality Sensitive Hashing for large scale graph learning — US63517869, Provisional - Optimizing training sets used for setting up inspection-related algorithms — US10267748, Granted ## Awards and honors - IEEE Rising Star in Signal Processing - ICASSP 2023 - IEEE Best Student Paper Award - ICIP 2020 - Ming-Hsieh Ph.D. Scholar Finalist 2022-23 ## Academic activities - Reviewer: IEEE Journals (JSAIT, TSIPN, SPL, TNNLS) - Reviewer: Conferences (ICASSP, ICLR, NeurIPS, LoG, ICML, COLM) - Area Chair: NeurIPS Download PDF: [Curriculum Vitae (PDF)](https://shekkizh.com/files/Sarath%20Shekkizhar%20CV.pdf) Canonical URL: https://shekkizh.com/cv