> ## Documentation Index
> Fetch the complete documentation index at: https://labs.laer.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Evaluation Workflow 1: Building Review Model Instructions with Generative Inquiry

An example workflow with AIDA is illustrated below:

1. Add a new issue to the existing review protocol
2. Perform additional research through AIDA’s generative inquiry to obtain more context
3. Revise the issue based on the findings
4. Rerun the predictive model

<Frame>
  <img src="https://mintcdn.com/epiqai/nS5UcVIcNVyp4Qxh/images/6.png?fit=max&auto=format&n=nS5UcVIcNVyp4Qxh&q=85&s=0123ba62e207d6d78b9c259037b6a39c" alt="" width="1736" height="275" data-path="images/6.png" />
</Frame>

&#x20;As an example, consider a new issue a user adds within the protocol.

**`Financial Integrity and Audit Oversight`**

**`Objective:`**` ``To investigate the integrity of financial reporting and the effectiveness of internal and external audit processes, including the relationship with the auditor.`

**`Evaluation Criteria:`**

* `Accuracy and transparency of financial statements and reports`
* `Evidence of Arthur Andersen’s audit practices and any conflicts of interest`
* `Records of internal audits and any action taken in response to findings`

Some of the follow-up questions the user can then ask through AIDA’s generative inquiry include:

* What is the working relationship between Enron and Andersen?
* What was the action taken by Andersen when it found out about the fraudulent acts of Enron?
* Which accounting firm audited Enron’s financial statements?
* What role did Arthur Andersen, Enron’s auditing firm, play in reporting Enron’s financial statements?