Hi amc and thanks for submitting the first question.
Without knowing the specifics of the questionnaire and the output, my initial assumption is that True Potential’s solution has likely digitised a simple process rather than using AI. A digitised process typically applies "if this, then that" logic to control the content of the output, such as the suitability report in this case. I assume the questions are basic, e.g., “Have your circumstances changed?” If the complexity increases beyond a certain point, it would likely be referred to an adviser.
The primary challenge lies not in the technical implementation but in defining the decision logic—understanding what each response signifies and building out the coded document to reflect these decisions. While AI could perform a similar function, it requires a significant dataset to analyse and extract historical patterns to provide answers for future scenarios. For AI to be effective, we need a robust, digitised dataset from which it can learn. For simpler scenarios, a decision tree outlining the advice process could serve as the foundation for automation.
Digitising the process offers greater control over the advice given, but comfort levels, historical data for training, and cost remain significant hurdles in AI adoption. Considerations around contracts, privacy, and data security are also important factors.
Iress’ Digital Advice product currently uses elements of digitised logic but continually evaluates its product strategy. AI in this space is more of a long-term strategy rather than a short-to-medium-term solution. While Xplan has the potential to incorporate such automated processes, leveraging AI effectively requires careful planning and significant data analysis, making it a path worth exploring.
There are parts of the process that can be achieved with current Xplan functionality. The review check-in in Client Portal can be configured with your questions. Once notified that the client has completed it and the results have been entered into Xplan, the RoA could be generated based on the client’s responses. I suggest discussing your desired scenario with your Relationship Manager to see what can be achieved now.
Iress Team
