Deep Research Agents for Supply Chain

Deep research agents represent a rapidly emerging frontier in supply chain AI, where large language models (LLMs) are coupled with planning, tool use, and retrieval capabilities to conduct sophisticated, multi-step research tasks autonomously. Unlike traditional question-answering systems, deep research agents can decompose complex queries—such as "assess the geopolitical exposure of our tier-2 semiconductor suppliers"—into sub-tasks, gather evidence from heterogeneous sources, synthesise findings, and produce structured, citation-backed reports.

In the supply chain context, deep research agents are particularly valuable for tasks that have traditionally required extensive human analyst effort: supplier due diligence, regulatory compliance monitoring, ESG and sustainability auditing, competitor intelligence, and risk scanning across multi-tier networks. By orchestrating web search, document analysis, database queries, and structured reasoning, these agents can dramatically compress the time needed to produce actionable intelligence—while maintaining traceability through explicit citation of evidence.

Research in this area spans foundational work on LLM-based agent architectures (such as ReAct, reflection, and plan-and-solve paradigms), retrieval-augmented generation (RAG), tool-use frameworks, and domain-specific adaptations for supply chain knowledge. Contributions from the Supply Chain AI Lab at the University of Cambridge, alongside broader advances in agentic AI from OpenAI, Anthropic, and Google DeepMind, are shaping the methodology and benchmarks for this emerging field.

We invite you to explore the curated collection of key publications below, offering a gateway into this fast-moving area of research.

List of Publications

1. Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K. and Cao, Y., 2023. ReAct: Synergizing reasoning and acting in language models. International Conference on Learning Representations (ICLR). [PDF]
2. Shinn, N., Cassano, F., Gopinath, A., Narasimhan, K. and Yao, S., 2023. Reflexion: Language agents with verbal reinforcement learning. Advances in Neural Information Processing Systems, 36. [PDF]
3. Schick, T., Dwivedi-Yu, J., Dessi, R., Raileanu, R., Lomeli, M., Hambro, E., Zettlemoyer, L., Cancedda, N. and Scialom, T., 2023. Toolformer: Language models can teach themselves to use tools. Advances in Neural Information Processing Systems, 36. [PDF]
4. Xu, L., Almahri, S., Mak, S. and Brintrup, A., 2024. Multi-agent systems and foundation models enable autonomous supply chains: Opportunities and challenges. IFAC-PapersOnLine, 58(19), pp.795-800. [PDF]
5. Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.T., Rocktäschel, T. and Riedel, S., 2020. Retrieval-augmented generation for knowledge-intensive NLP tasks. Advances in Neural Information Processing Systems, 33, pp.9459-9474. [PDF]
6. Wang, L., Ma, C., Feng, X., Zhang, Z., Yang, H., Zhang, J., Chen, Z., Tang, J., Chen, X., Lin, Y. and Zhao, W.X., 2024. A survey on large language model based autonomous agents. Frontiers of Computer Science, 18(6), p.186345. [PDF]
7. Park, J.S., O'Brien, J.C., Cai, C.J., Morris, M.R., Liang, P. and Bernstein, M.S., 2023. Generative agents: Interactive simulacra of human behavior. Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology, pp.1-22. [PDF]