At Pirani, we have created Pirani Copilot, an advanced artificial intelligence that facilitates the identification, assessment, and agile management of operational risks
Our AI assistant is the ideal ally for organizations looking to optimize their processes and improve strategic decision-making at every stage of the risk management cycle.
Currently, our AI-powered risk management assistant is available for the following modules: Operational Risk Management (ORM), Compliance Risk Management, Information Security Management System (ISMS), and Anti-Money Laundering (AML) Risk Management.
Guide to start using Pirani Copilot
Access the Operational Risk Management (ORM) module
Log in to the Pirani software and go to the ORM module. Once there, you’ll find an icon for our AI-powered risk management assistant in the top right corner.
Interact with our risk management assistant.
When you start a conversation, Copilot will suggest two initial questions to help understand the risk context. You can answer them or ask your own questions about risks..
How does Pirani Copilot suggest risks?
Copilot analyzes risks based on the information provided during the onboarding process. This includes: industry or sector, organization name, and organization size.
The assistant performs a contextual reading of this data to suggest relevant risks.
To ensure accurate suggestions, go to "Organization Administration" > "Organization Settings" > Other Data, and complete the required information so that Copilot can generate more detailed risk suggestions.
Generate and manage risks
Pirani Copilot suggests up to 5 risks per interaction, allowing you to reach a maximum of 15 risks per session. Once this limit is reached, the AI risk assistant will ask you additional questions to help you refine and prioritize risks based on your organization’s needs.
How are risks created by the user?
Once the risks have been generated, you can request them by name or by indicator. Copilot will ask for confirmation before proceeding with the creation.
Once confirmed, the risks are automatically generated in the “Risks” module.
Even if you're in another module, Copilot offers a redirection feature so you can go directly to the Risks module.
Classification and rating adjustment
Review the initial low rating of each risk and adjust it according to the level of impact and relevance within your organization’s context. This will help prioritize and define effective preventive and corrective actions.
When creating or editing a risk, the "Suggest Risk" option is now available.
Go to "Create" and select "Suggest a risk."
Select the industry and the corresponding process. (Optional) Add a detailed description—the more information you provide, the more tailored the suggested risks will be.
Then, click “Suggest risks.”
The system will propose between 5 and 7 risks adjusted to the context you provided.
When you select one of the risks, the associated fields are automatically filled in.
Don't want to use Copilot in your organization?
Go to “Manage Organization” > “Security”.
In the “Advanced Options” section, you can enable or disable the AI functionality for the entire organization.
Frequently Asked Questions
What is the information security in Pirani Copilot?
The Amazon-based virtual assistant Bedrock has been designed to prioritize information security. Its training is conducted exclusively with public documents, such as regulations and operational risk management studies, and is complemented by pre-training on Bedrock's base models. Client-supplied data is used only to generate specific responses during each interaction; it is not stored or used to retrain the model.
Furthermore, being built on AWS infrastructure, the service takes advantage of AWS's high-security standards, such as data encryption in transit and at rest, continuous auditing, and strict access controls. This ensures information protection and gives users the peace of mind they need.
Can I modify the questions suggested by the AI?
Yes. While the interaction begins with two suggested questions, you are free to ask your own to tailor the risk analysis to your specific context.
How many risks can I create in each session?
Each interaction gives you up to 5 risks, with a total limit of 15 per conversation with our AI agent. Once this limit is reached, you will ask additional questions to refine the generated risks.
Can I adjust the rating of the generated risks?
Yes, the risks receive a low initial rating, but you should modify this rating according to your business's criticality and context.