Our AI review assesses your controls and provides feedback fast
Purpose
The primary goal of our AI review program is to expedite the audit process, allowing for rapid progression from review to compliance. By providing immediate feedback on each control (pass/incomplete), the program pinpoints specific issues, enabling you to quickly address any shortcomings. It also surfaces key information to our audit team, that helps speed up the audit process. This not only ensures a smoother audit but also reduces the duration and complexity of achieving compliance.
Step 1: Running the AI Audit Models
Once you have provided us access, as outlined in the Quickstart Guide, we proceed by exporting your controls evidence package or accessing it directly through the API. Your compliance documents and data are then processed through our proprietary product, Lexi, which leverages the OpenAI API with enterprise security applied, to assess each control against the defined audit criteria. The results are categorised as pass or incomplete for each control, which form the basis for subsequent steps.
Step 2: Human and AI Collaborative Audit
To enhance the reliability of our AI findings, a human-audited review is conducted in parallel with the AI assessment. This step allows us to validate the AI-generated results and further refine the outcomes by addressing any discrepancies identified by our expert auditors. This collaborative approach not only minimizes errors (reducing noise) but also confirms controls that may have been initially marked incomplete by the AI review, further expediting the compliance process with a higher initial pass rate of the controls. The validated results, along with a detailed list of queries for unresolved items, are then shared with you.
Step 3: Resolution and Finalization
The final step mirrors the traditional audit approach, where you address the queries raised in the previous step. This phase involves resolving outstanding issues based on the feedback provided, which moves the audit towards completion. Our team supports throughout this process, ensuring that audit queries are adequately addressed before finalising the audit.