Jayaprakash Sundararaj

I am a Senior Software Engineer at Google USA, working in search ranking and personalization projects. My work focuses on developing and training deep learning models and leveraging large language models (LLMs) for search ranking. Notably, I launched: [1] Fact checking articles in Google News and Google Web Search, [2] On-Device Univeral Search in Pixel System Intelligence team, and [3] Query recommendations in Play Store Search Home.

I have a post-graduation degree in CSE from IIT Bombay, India & undergraduate degree in IT from Anna University, Chennai. Previously, I have worked at Yahoo! Research and Alcatel Lucent Labs.

Interests

I enjoy working on applied research problems and building usable projects that leverage state-of-the-art machine learning models. Lately, I’m focused on pursuing the interpretability and explainability aspects of machine learning research.

Projects

[Google] Query recommendations in Play Store Search Home
- As part of the Google Play Search team, I contributed to developing query recommendation algorithms aimed at improving user engagement and discovery within the Play Store. This project focused on leveraging user behavior data and query semantics to provide personalized, high-quality query suggestions.
[Google] On-Device Universal Search in Pixel System Intelligence Team
- While working with the Pixel System Intelligence team, I helped design and implement a universal search system for (more than 10 million) Pixel devices. This system provided seamless, on-device content retrieval across apps and services, ensuring speed, privacy, and efficiency.
[Google] Fact checking articles in Google news and web search
- I worked on enhancing Google’s fact-checking capabilities by integrating article-level fact-checking summaries into Google News and Web Search. This involved developing algorithms to identify and highlight fact-checking information, improving the reliability of information surfaced to users.
- This project made a significant impact in both industry and research, generating 4 billion impressions annually. It has contributed to over 300 research papers published in prestigious conferences and journals such as NeurIPS, ACL, and EMNLP. Additionally, the tool has been referenced more than 800 times across various GitHub projects.
[Google] Query understanding in web search
- I contributed to improving query understanding in Google’s Web Search team. My work focused on leveraging natural language processing and semantic analysis to better interpret user intent, ultimately improving search result relevance.
[Yahoo] Salient terms extrations for Image articles
- During my internship as a researcher at Yahoo! Research Labs, I worked on extracting salient terms from image articles as part of a broader effort in content understanding. These extracted salient terms were subsequently utilized for retrieval and scoring during serving time, enhancing the relevance and accuracy of content-based operations.
- Following the successful completion of the internship, the position transitioned into a year-long academic collaboration between Yahoo! Research Labs and IIT Bombay. As part of this collaboration, I was awarded a research grant of INR 1,00,000. You can learn more about this work from the presentation.
[IITB] Document Summarization
- At IIT Bombay, I worked on automatic document summarization using machine learning and natural language processing techniques. This project aimed to generate concise and accurate summaries of large text documents, aiding quick information retrieval.
[IITB] Statistical Machine Translation
- I was involved in research on statistical machine translation at IIT Bombay. The focus was on developing and improving translation models for Indian languages, utilizing statistical techniques to enhance accuracy and fluency in translations.

Publications

[2024] J Sundararaj, A Vyas, B Gonzalez-Maldonado, Automated LaTeX Code Generation from Handwritten Math Expressions Using Vision Transformer, arXiv preprint arXiv:2412.03853, 2024. https://arxiv.org/abs/2412.03853
[2024] J Kumar, K Jayakumar, J Sundararaj, ESMCrystal: Enhancing Protein Crystallization Prediction Through Protein Embeddings, Proceedings of CIBB, 2024. https://easychair.org/publications/preprint/FTCX/
[2015] J Jayanth, J Sundararaj, P Bhattacharyya, Monotone submodularity in opinion summaries, Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing. https://doi.org/10.18653/v1/D15-1017
[2015] J Sundararaj, J Jayanth, P Bhattacharyya, Opinion Summarization using Submodular Functions: Subjectivity vs Relevance trade-off, 16th International Conference on Intelligent Text Processing and Computational Linguistics. Best verifiability, reproducibility, and working description award.
[2014] J Sundararaj, P Bhattacharyya, Document Summarization with applications to Keyword extraction and Image Retrieval, Indian Institute of Technology, Bombay
[2013] J Sundararaj, Survey on Semantic Web and Search Techniques, Indian Institute of Technology, Bombay
[2013] J Sundararaj, Technical Paper on Kappa Score - Interrater Annotation Agreement, Indian Institute of Technology, Bombay

Misc

Received over 20 Google Awards (Peer and Spot bonus) for my contributions.
Reviewer for ACL (2024, 2025), EMNLP (2024), IEEE TNNLS (2023, 2024).
Awarded as Search Professional of the Year 2024 by BCS Information Retrieval Specialist Group.
Elected as BCS Fellow member.
Awarded as Distinguished Alumni from Sona College of Technology, India (affiliated with Anna University, Chennai).
Teaching assistant for the courses in IIT Bombay:
- Advanced Distributed Systems
- Machine Learning
- Artificial Intelligence
Organized COLING (Computational Linguistics 2012) International Conference held at IIT Bombay.
Won over 20+ Programming competitions held in international technical symposiums (includes Bits Pilani Goa, Anna University Chennai).
Received The Best Outgoing Student Award during undergraduate studies.
Received Leader Altogether Award from IT Department during undergraduate studies.

Talks

TBD, Futurist Expo, 21-23 March 2025
Getting started with LLM applications at SlavConf Envision, 26-27 July 2024.
Delivered a technical talk on Kappa Score - Inter Annotator Agreement at IIT Bombay.
Presented a seminar on Semantic Web at IIT Bombay.