AI-Powered IT Project Risk Management System Using Multi-Agent Architecture, RAG, and LangGraph

References

[1] Project Management Institute: A Guide to the Project Management Body of Knowledge (PMBOK Guide), 7th edn. PMI, Newtown Square, PA (2021).

[2] International Organization for Standardization: ISO 31000:2018 Risk Management – Guidelines. ISO, Geneva (2018).

[3] Boehm, B.W.: A spiral model of software development and enhancement. Computer 21(5), 61–72 (1988).

[4] Shrestha, A., Cater-Steel, A., Toleman, M., Rout, T.: A systematic literature review of IT project risk management. Inf. Softw. Technol. 97, 1–21 (2018).

[5] Elish, M.O.: Predicting defect-prone software modules using support vector machines. J. Syst. Softw. 81(5), 649–660 (2008).

[6] Brown, T.B., et al.: Language models are few-shot learners. In: Advances in Neural Information Processing Systems, vol. 33, pp. 1877–1901 (2020).

[7] Lewis, P., et al.: Retrieval-augmented generation for knowledge-intensive NLP tasks. In: Advances in Neural Information Processing Systems, vol. 33, pp. 9459–9474 (2020).

[8] Chase, H.: LangChain: Building applications with LLMs through composability. GitHub Repository, https://github.com/langchain-ai/langchain (2022).

[9] LangGraph Documentation: Building stateful multi-agent applications with LangGraph. LangChain Inc., https://langchain-ai.github.io/langgraph/ (2023).

[10] Chroma: The AI-native open-source embedding database. Trychroma, https://www.trychroma.com/ (2023).

[11] Touvron, H., et al.: Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288 (2023).

[12] Jiang, A.Q., et al.: Mistral 7B. arXiv preprint arXiv:2310.06825 (2023).

[13] Wu, Q., et al.: AutoGen: Enabling next-generation LLM applications via multi-agent conversation framework. arXiv preprint arXiv:2308.08155 (2023).

[14] Groq Inc.: Groq API documentation. https://console.groq.com/docs/openai (2024).

[15] Streamlit Inc.: Streamlit – the fastest way to build and share data apps. https://streamlit.io/ (2024).

[16] Changder, S.: AI-Powered IT Project Risk Management System. GitHub Repository, https://github.com/soumadipchangder/AI-Powered-IT-Project-Risk-Management-System (2024).

[17] Standish Group: Chaos Report 2020. The Standish Group International Inc. (2020).

[18] Raza, S., Nadeem, M., Iqbal, R.: Machine learning approaches for software project risk assessment: A systematic review. IEEE Access 9, 120453–120470 (2021).

@article{1716719,
      author = {Soumyadip Changder, Pinaki Karmakar},
      title = {AI-Powered IT Project Risk Management System Using Multi-Agent Architecture, RAG, and LangGraph},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {9},
      number = {10},
      pages = {2376-2383},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1716719.pdf},
      abstract = {Effective risk management remains one of the most persistent challenges in information technology project delivery. Traditional approaches rely heavily on periodic manual assessments, static checklists, and subject-matter intuition, which collectively fail to keep pace with the dynamic and interconnected nature of modern software projects. This paper presents an AI-Powered IT Project Risk Management System that addresses these limitations through a coordinated multi-agent architecture orchestrated via LangGraph, augmented with Retrieval-Augmented Generation (RAG) backed by ChromaDB, and driven by large language models accessed through the Groq API. The system comprises four specialised agents—a Market Analysis Agent, a Risk Scoring Agent, a Project Tracking Agent, and a Reporting Agent—that operate in a defined pipeline to evaluate both exogenous market signals and endogenous operational indicators. Risk dimensions including market, technical, financial, regulatory, and operational factors are individually scored on a 0–100 scale and consolidated into a structured JSON report surfaced through an interactive Streamlit dashboard. Empirical evaluation on a representative ERP implementation scenario yields an overall risk score of 66/100 (High) with a 68 % schedule-delay probability, demonstrating the system’s capacity to produce actionable, prioritised mitigation guidance. The architecture is designed for extensibility and real-world deployment, with future work targeting live Jira integration, reinforcement-learning-based adaptive scoring, and mobile-accessible reporting interfaces.},
      keywords = {Risk Management, Multi-Agent Systems, Large Language Models, Retrieval-Augmented Generation, LangGraph, LLM Orchestration, IT Project Management},
      month = {April},
      doi = {https://doi.org/10.64388/IREV9I10-1716719}
  }