Research initiative
Mission-critical operational intelligence for real-time enterprise decisions where accuracy, reliability, and explainability matter.
Mission-critical intelligence
RODI is designed for mission-critical operational intelligence, where decisions must be accurate, reliable, and explainable. In high-impact enterprise and infrastructure applications, AI systems cannot rely on plausible but uncertain outputs, and hallucination is not acceptable.
Our mission is to build a new class of real-time, on-demand intelligence systems that combine the flexibility of modern AI with the reliability required for real-world operations.
The vision of RODI is to move beyond general-purpose AI assistants toward systems that can reason, verify, adapt, and support decisions in complex environments such as supply chains, industrial planning, transportation, trade services, and production workflows.
RODI brings together large language models, formal systems, causal inference, non-monotonic logic, and category-theoretic perspectives to study how intelligent systems can reason under uncertainty while still maintaining correctness.
RODI connects LLM-driven interaction, structured workflows, formal verification, causal models, and real-time operational data to support decisions that can be acted on in practice. Current examples include Fashion Flow, Industry Optimization, and Smart Trade AI, and we welcome collaboration with partners who want to apply RODI technologies in other mission-critical domains.
System architecture
The RODI system connects real-time enterprise data, operational constraints, and task requests with grounded state modeling, LLM-based planning, structured knowledge, formal verification, causal inference, non-monotonic logic, category-theoretic abstractions, and an internal self-evolution loop to produce reliable decision support and deployable application adapters.
Demos
Fashion Flow connects fashion design intelligence with real production execution. It supports clothing design suggestions, trend-aware product concepts, and the coordination of factories, garment resources, capacity, production, and distribution as a product-ready AI workflow for the fashion industry.
VideoCompetitor Analysis is a project for comprehensive market analysis that draws on multi-source information—web research, platform data, product signals, and structured reporting—to map competitive landscapes, compare offerings, and support strategic decisions with evidence-backed insights.
Open demoIndustry Optimization focuses on AI-supported industrial and transportation optimization, improving resource allocation, production coordination, logistics decisions, transportation workflows, and real-time enterprise operations for deployable commercial and industrial applications.
Smart Trade AI is an AI-powered marketplace framework for skilled trade businesses, using LLMs, task templates, quotation intelligence, matching, scheduling, compliance checks, and feedback learning to support a commercially viable trade-service platform.
Read documentRODI can be adapted to additional domains such as real estate, asset management, infrastructure services, enterprise operations, and other settings where reliable real-time intelligence can improve decisions, workflows, and commercial outcomes.
Discuss collaborationMembers
Our team brings together research backgrounds in adaptive reasoning, privacy-preserving AI, sustainable infrastructure, enterprise modeling, and intelligent operational systems.
Principal Investigator
Dr. Jie Wang is the Executive Director of the Stanford Center for Sustainable Development and Global Competitiveness. His research spans adaptive computational learning, enterprise modeling, sustainable development, smart infrastructure, knowledge systems, decision frameworks, smart manufacturing, and business innovation. He has conducted research at Stanford, advised Silicon Valley startups, consulted for multinational companies and government agencies, and served as a senior advisor to venture-backed startups.
Postdoctoral Researcher
Dr. Jiechao Gao received his Ph.D. in Computer Science from the University of Virginia and M.S. in Electrical Engineering from Columbia University, and previously served as an Associate Research Scientist at Columbia University. His research spans federated learning, LLM interpretability and efficiency, reinforcement learning, and privacy-preserving AI, with applications in healthcare, smart buildings, IoV, and finance. He has published at major venues including ICLR, NeurIPS, ICML, AAAI, EMNLP, KDD, CVPR, and ACL, and was named a Stanford/Elsevier Global Top 2% Scientist in 2024 and 2025.
Postdoctoral Researcher
Dr. Yuandong Pan is a Postdoctoral Researcher in Stanford School of Engineering and was previously a Marie Sklodowska-Curie Actions Future Road Research Fellow at the University of Cambridge. His research develops digital and smart approaches for sustainable buildings, infrastructure, and cities, improving how the built environment is designed, managed, and maintained to support smarter, more resilient, and more sustainable urban systems.
Publications
Selected publications from 2023 onward, with RODI team members highlighted in the author lists.
Yuandong Pan, M Wang, Jiechao Gao, Jie Wang, MD Lepech, I Brilakis
Automation in ConstructionJiechao Gao, RK Yadav, Y Li, Yuandong Pan, Jie Wang, Y Liu, M Lepech
ACL 2026Yuandong Pan, M Wang, L Lu, R Lamsal, E Parn, S Zlatanova, I Brilakis
Automation in ConstructionJ Liu, T Jin, Yuandong Pan, Y Li
Journal of Construction Engineering and ManagementY Xie, S Li, L Lu, Yuandong Pan, F Iida
Expert Systems with ApplicationsD Koo, Jiechao Gao, Yuandong Pan, Jie Wang, M Lepech
SocialLLM@ICWSM 2026Y Yao, Y Yao, X Fan, Jiechao Gao, Jie Wang, M Zhang, S Ravi, C Joe-Wong
KDD Blue Sky Track 2026Zixiong Yang, Linxiao Li, Jiaye Lin, Binrui Wu, Xiaoyu Kang, Jiechao Gao
ICML 2026Jie Wu, Haoran Ma, Shisong Tang, Yulin Xu, Xiaoyu Kang, Jiechao Gao
ICML 2026Siyuan Chen, Ding Hang, Xiaoyu Kang, Jiechao Gao
ICLR 2026Jiechao Gao, C Liu, Y Li
AAAI 2026Y Li, Y Shen, Y Nian, Jiechao Gao, Z Wang, C Yu, S Li, Jie Wang, X Hu, Y Zhao
AAAI 2026Abdullah Mohammad, Sushant Kumar Ray, Pushkar Arora, Rafiq Ali, Ebad Shabbir, Gautam Siddharth Kashyap, Jiechao Gao, Usman Naseem
ACL Industry Track 2026Jiechao Gao, RK Yadav, X Huang, Jie Wang
IEEE BigData 2025X Huang, Jiechao Gao, Jie Wang
IEEE BigData 2025Hanwen Hu, Cheng Zeng, Shiyou Qian, Jianhua Zhou, Jian Cao, Yirong Chen, Jie Wang, Han Han
IEEE T-ITS 2025L Lu, AM d'Avigneau, Yuandong Pan, Z Sun, P Luo, I Brilakis
Automation in ConstructionYuandong Pan, L Lu, M Wang, VK Reja, F Noichl, A Borrmann, I Brilakis
Computing in Civil Engineering 2025T Wang, J Liu, Yuandong Pan, H Li
Journal of Management in EngineeringS Zhou, JR Lin, P Pan, Yuandong Pan, I Brilakis
Automation in ConstructionC Feng, L Huang, K Wang, Y Cao, M Zhu, Jiechao Gao, XY Liu
NeurIPS Workshop 2025Y Wu, S Chen, B Wu, F Li, Jiechao Gao
NeurIPS 2025C Cui, S Tang, F Li, Jiechao Gao, H Chen
ICML 2025B Wu, S Tang, F Li, B Han, C Meng, J Xiao, Jiechao Gao
KDD 2025B Wu, H Sui, J Lin, Jiechao Gao, T Xu, K Jin, X Zhang
ACM MM 2025Yuandong Pan, M Wang, L Lu, R Wei, S Cavazzi, M Peck, I Brilakis
Automation in ConstructionL Lu, Yuandong Pan, I Brilakis
Computing in Civil Engineering 2024XY Liu, R Zhu, D Zha, Jiechao Gao, S Zhong, M Qiu
ACM TMIS 2024Hanwen Hu, Shiyou Qian, Jingchao Ouyang, Jian Cao, Han Han, Jie Wang, Yirong Chen
IEEE T-ITS 2023Yuandong Pan, A Braun, A Borrmann, I Brilakis
Smart Infrastructure and ConstructionJiechao Gao, W Wang, F Nikseresht, V Govinda Rajan, B Campbell
ICPP 2023Contact
Contact the RODI team to discuss research collaboration, demos, project documents, or enterprise operational intelligence applications.