We are excited to release the powerful integration between our Cloverpop Decision Intelligence Platform and OpenAI’s ChatGPT generative AI! Six months in the making, this new feature acts as a brainstorming assistant to help decision-makers structure decisions quickly and avoid blind spots by suggesting impactful drivers and KPIs to improve decision quality.
Our Decision Success team will contact our current customers to enable the functionality for their company, and prospective customers contact our sales team today for a demo of this powerful functionality.
Using ChatGPT Generative AI For A Broader Decision-Making Perspective
The new feature integrates with Cloverpop’s decision tree capabilities to help create new decision trees or improve existing ones by building on decision makers’ thoughts to suggest additional decision drivers and KPIs.
To use the feature when creating a new decision tree in Cloverpop, decision-makers first describe the decision and at least two important decision drivers to give context for additional driver suggestions from ChatGPT.
Then our secure, private integration with ChatGPT returns suggestions for possible additional drivers, which can be edited by the decision maker and added to the decision tree.
Decision makers can also get suggestions for key performance indicators (KPIs) that support individual drivers and make any needed edits before adding them to the decision tree.
These capabilities are also available when reviewing and improving existing decision trees, including decisions created from decision templates.
A Thoughtful Approach To Designing A Generative AI Decision Brainstorming Assistant
While ChatGPT incorporates a wide range of human knowledge, we took great care in creating our integration to overcome its limited ability to analyze specific decisions, increase the accuracy of its suggestions and limit the known risk of inaccurate suggestions while maintaining customer data security. Ultimately, we designed it as a brainstorming assistant to help decision-makers think more broadly when framing their decisions.
As such, our integration has four core principles to act as a secure brainstorming thought partner:
The human decision-maker is the expert. Decision makers must frame the decision and describe at least two critical drivers before any suggestions are generated.
The human decision-maker is in charge. The decision maker chooses which suggestions are most helpful, and the suggestions can all be edited before being added to the decision tree.
The ChatGPT suggestions incorporate human decision-maker input. As the decision maker edits and adds drivers and KPIs to the decision tree, Chat GPT iteratively incorporates that new context into additional suggestions for that tree.
The ChatGPT integration is secure and private. All interactions with ChatGPT through the Cloverpop interface are via secure API, and none of the content is shared with or incorporated into the ChatGPT language model.
This brainstorming assistant approach is not an expert recommendation engine – the human decision-maker is still the expert in charge! However, the experience is fast and easy to understand and provides a powerful tool for more expansive thinking when faced with complex and important decisions. We can’t wait to see how it helps our customers brainstorm better, faster decisions.