The work done in KPMG Origins often existed within complex supply/value chain environments, with a diverse and often obscure list of stakeholders, materials and agreements. Capturing who is involved, what they do, and how they interact with each other was often a challenge. Often manual research of verified media and government articles, market reports and ASIC Reports by the data and analytics team identified significant examples of counter-party risk within these ecosystems. The team was able to successfully replicate this manual investigation using a web-scraping, Generative AI and RAG approach and layer the signals and summarised anomaly messages into other modelling and visualisation work to create meaningful and actionable insights. We also successfully extended this to detect anomalous behaviour from contextual clues in the raw Origins data. Using the summaries and signals from other models, for instance the Material Flow Analysis of hazardous waste movements in NSW, were fed into a GPT-style model. Targeted prompt engineering led to several successful real-world and otherwise unidentified suspect movements being identified and relayed for further regulator investigation.