Overview
This feature displays a text box on the results details page for each Disclosure, Breach, and Warning, offering context based on how the user has interacted with similar results in the past.
This is designed to reduce the time spent manually researching past instances for the same issuer. Guided by client feedback, future updates will provide even deeper insights into exactly why a result was triggered, and we’re eager for your ongoing input to help us shape these next steps.
FAQ
Q: What model is being used?
A: The Result Insights feature uses Anthropic’s Claude Haiku 4.5.
Q: Can I select the model I want to use?
A: No, given the complexity and the guardrails that need to be put in place, you have to rely on the model selected by FundApps.
Q: What guardrails do you have in place?
A: The models we use are hosted on AWS Bedrock. AWS provides native industry-leading guardrail solutions. Specifically, we use:
Content filters (help detect and filter harmful user inputs and model-generated outputs)
Word filters (namely, the profanity filter to block profane words)
We mask any personal information (names, email addresses)
Q: Is client data used to train the model?
A: No, client data is not used to train the model powering the results insights.
Q: Can I choose not to use Results Insights?
A: Yes. Results Insights is an opt-in feature; you can choose whether to use it. Additionally, approval can be withdrawn at any point in time.