video1.0<iframe src="https://www.loom.com/embed/b252056534fa4c6aaf08d450f954bd57" frameborder="0" width="1660" height="1245" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12451660Loomhttps://www.loom.com12451660https://cdn.loom.com/sessions/thumbnails/b252056534fa4c6aaf08d450f954bd57-3648afcd831684ba.gif101.919This Loom introduces Audit Trace, an application that analyzes uploaded documents for potential fraud using an LLM workflow. It explains the stages from ingesting and pre-analyzing, including creating global documents used across the fraud analysis phases. The presenter then shows an example report where auditors review findings such as a 3 million bank transfer after hours without approval, with descriptions, source links, likelihood percentages, and an AI chat to ask why an item is flagged. After human approval in the loop, the report updates results including cash misappropriation and the corrected profit totals for the company.