Exciton Research

Team

Tian Shi

Tian Shi

Founder, Exciton Research

Tian Shi is the founder of Exciton Research, LLC. His work spans machine learning research, retrieval, and the design and operation of production AI systems. Path to AI Architect is a recent synthesis of that experience.

Tian is the founder and sole engineer of RauGen, a multimodal AI studio for image, video, audio, and music generation, and was CTO of Neutriva, an AI and ecommerce platform. Before that, Tian founded BizLidar, a retrieval and machine learning platform over 800 million news documents, and was Director of Machine Learning at Moody's Analytics.

Tian holds PhDs in computer science from Virginia Tech and computational chemistry from Wayne State University, and has published 22 papers at venues including AAAI, The Web Conference, CIKM, NAACL, IEEE TNNLS, IEEE TKDE, and ACM TKDD.

GitHubRauGenContact

Books

Experience

  1. Founder and Lead Engineer, RauGen

    Feb 2025 – Aug 2025, Jun 2026 – present

    • Sole engineer of a multimodal AI studio for image, video, audio, and music generation on hosted models from Gemini, OpenAI, Claude, and Fal.
    • Next.js and TypeScript web app, a FastAPI service behind a metered developer API, and a worker queue with FFmpeg for transcoding and music-video assembly.
    • Object storage and a streaming CDN for assets; per-call usage metering tied to Stripe billing.
    • Owns model selection, latency, inference cost, and reliability.

  2. Chief Technology Officer, Neutriva

    Aug 2025 – Jun 2026

    • Architected and launched an integrated AI, content, and ecommerce platform.
    • Built a wellness question-answering system with personalized product recommendations, and an in-product assistant.
    • Built authentication, payments, and order management.

  3. Founder and Founding Engineer, BizLidar

    Apr 2024 – Feb 2025

    • Data platform that ingests and enriches more than 800 million news articles and blogs.
    • Index of 20 million business and risk articles and 300,000 company profiles covering adverse media, fundraising, and M&A.
    • Retrieval-augmented generation over Elasticsearch, with language models translating plain-language questions into search queries.
    • Classifiers for business signals such as funding events, multilingual named-entity recognition, and infrastructure to fine-tune and serve open-source models.

  4. Director of Machine Learning, Moody's Analytics

    Sep 2021 – Feb 2023

    • Production models on large news streams for ESG controversy screening.
    • Event detection for commercial real estate and infrastructure from unstructured news.
    • Many-to-many neural machine translation service for global news ingestion.
    • Prototyped GPT-3 report generation from structured data for analysts.

Open-source software

  • LeafNATS: Open-source toolkit for neural abstractive summarization and related NLP tasks, demonstrated at NAACL 2019.
  • GPView: Wavefunction analysis and visualization for excited states of molecules, in over 50,000 lines of C++.

Education

  • PhD, Computer Science, Virginia Tech

    2016 – 2021

    Advisor: Chandan K. Reddy. Dissertation: Novel Algorithms for Understanding Online Reviews.

  • PhD, Computational Chemistry, Wayne State University

    2011 – 2016

    Advisor: Vladimir Y. Chernyak. Dissertation on how geometric distortions scatter electronic excitations in conjugated macromolecules.

Research

  • Neural abstractive summarization with sequence-to-sequence models and reinforcement learning
  • Self-supervised aspect detection and multi-aspect rating prediction for reviews
  • Interpretable, concept-based document classification
  • Text-to-SQL and question answering over medical records and knowledge bases
  • Natural-language querying of NoSQL databases
  • Short-text topic modeling
  • Excited states and energy transfer in conjugated molecules

Publications

  1. Wenlong Zhang, Tian Shi, Ping Wang. Natural Language Querying on NoSQL Databases: Opportunities and Challenges (Vision Paper). IEEE BigData, 2024.
  2. Wenlong Zhang, Chengyang He, Guanqun Yang, Dipankar Bandyopadhyay, Tian Shi, Ping Wang. Natural Language Querying on Domain-Specific NoSQL Databases with Large Language Models. BIBM, 2024.
  3. Tianyi Li, Ping Wang, Tian Shi, Yali Bian, Andy Esakia. Task as Context: A Sensemaking Perspective on Annotating Inter-Dependent Event Attributes with Non-Experts. HCOMP, 2023.
  4. Wenlong Zhang, Kangping Zeng, Xinming Yang, Tian Shi, Ping Wang. Text-to-ESQ: A Two-Stage Controllable Approach for Efficient Retrieval of Vaccine Adverse Events from NoSQL Databases. BCB, 2023.
  5. Wenlong Zhang, Bhagyashree Ingale, Hamza Shabir, Tianyi Li, Tian Shi, Ping Wang. Event Detection Explorer: An Interactive Tool for Event Detection Exploration. ACM IUI, 2023.
  6. Tianyi Li, Ping Wang, Tian Shi, Andrey Esakia. Exploring the Impact of Sub-Task Inter-Dependency on Crowdsourced Event Annotation. HCOMP, 2022.
  7. Ping Wang, Tian Shi, Khushbu Agarwal, Sutanay Choudhury, Chandan K. Reddy. Attention-based Aspect Reasoning for Knowledge Base Question Answering on Clinical Notes. ACM BCB, 2022.
  8. Tian Shi, Xuchao Zhang, Ping Wang, Chandan K. Reddy. Corpus-Level and Concept-Based Explanations for Interpretable Document Classification. ACM TKDD 16(3), 2022.
  9. Tian Shi, Liuqing Li, Ping Wang, Chandan K. Reddy. A Simple and Effective Self-Supervised Contrastive Learning Framework for Aspect Detection. AAAI, 2021.
  10. Tian Shi, Ping Wang, Chandan K. Reddy. Deliberate Self-Attention Network with Uncertainty Estimation for Multi-Aspect Review Rating Prediction. arXiv:2009.09112, 2020.
  11. Tian Shi, Yaser Keneshloo, Naren Ramakrishnan, Chandan K. Reddy. Neural Abstractive Summarization with Sequence-to-Sequence Models. IEEE TNNLS, 2020.
  12. Ping Wang, Tian Shi, Chandan K. Reddy. A Tensor-Based Multi-Task Survival Analysis Model. IEEE TKDE, 2020.
  13. Ping Wang, Tian Shi, Chandan K. Reddy. Text-to-SQL Generation for Question Answering on Electronic Medical Records. The Web Conference (WWW), 2020.
  14. Yaser Keneshloo, Tian Shi, Chandan K. Reddy, Naren Ramakrishnan. Deep Reinforcement Learning for Sequence-to-Sequence Models. IEEE TNNLS, 2019.
  15. Tian Shi, Vineeth Rakesh, Suhang Wang, Chandan K. Reddy. Document-Level Multi-Aspect Sentiment Classification for Online Reviews of Medical Experts. CIKM, 2019.
  16. Tian Shi, Ping Wang, Chandan K. Reddy. LeafNATS: An Open-Source Toolkit and Live Demo System for Neural Abstractive Text Summarization. NAACL (Demonstrations), 2019.
  17. Tian Shi, Kyeongpil Kang, Jaegul Choo, Chandan K. Reddy. Short-Text Topic Modeling via Non-Negative Matrix Factorization Enriched with Local Word-Context Correlations. The Web Conference (WWW), 2018.
  18. Tian Shi, Ping Wang. GPView: A Program for Wavefunction Analysis and Visualization. Journal of Molecular Graphics and Modelling 70, 2016.
  19. Michael J. Catanzaro, Tian Shi, Sergei Tretiak, Vladimir Y. Chernyak. Counting the Number of Excited States in Organic Semiconductor Systems Using Topology. J. Chem. Phys. 142(8), 2015.
  20. Tian Shi, Hao Li, Sergei Tretiak, Vladimir Y. Chernyak. How Geometric Distortions Scatter Electronic Excitations in Conjugated Macromolecules. J. Phys. Chem. Lett. 5(22), 2014.
  21. Songnian He, Tian Shi. Strong Convergence Theorems by Generalized CQ Method in Hilbert Spaces. Fixed Point Theory 12(2), 2011.
  22. Tian Shi, Songnian He. Modified Hybrid Algorithms for Lipschitz Quasi-Pseudo-Contractive Mappings in Hilbert Spaces. Computers & Mathematics with Applications 59(8), 2010.