GenAI and the IRP: Accelerating insights for risk decision makers
By Moe Khosravy, executive vice president of Engineering, Moody’s RMS
Even after the successes in delivering the Intelligent Risk Platform (IRP) and a suite of applications, we innovate further to boost user productivity, and generative AI – or GenAI for short – comes at an opportune time to potentially transform business productivity and reduce time to insights.
For example, if you use ChatGPT or other large language models, you may feel today’s search experience is archaic and time-consuming when GenAI does everything for you. Consumer demands for search productivity are forcing a GenAI transition. So, should risk professionals demand ‘leapfrog’ productivity gains from today’s solutions?
At our Exceedance conference in New York, we used GenAI integrated with the IRP. Using Risk Data Lake data, within seconds answers to questions including “what is my total insured value exposure to events similar to Hurricane Ida?” and “how many properties in total would be impacted, in which state?” were delivered.
We avoid untrusted, hallucinated data by not using public APIs or exposing sensitive data. Instead, we use several state-of-the-art GenAI patterns, conditioning the interaction between users and a secure data interchange locked within our solution. Feedback from clients has been incredible as we iterate on this.
We are demonstrating another concept featuring GenAI: using a bot to accelerate data to insights, and insights to actions, we generate answers to plain-language questions and produce reports, charts and visualisations directly from the platform data. The results have been very encouraging! With an event approaching, imagine being able to ask: “Write a 500-word email in a formal tone summarising top-10 loss drivers, with a chart of…,” all without ever leaving the platform.
We’d love your feedback to evolve it – email RiskLabs@RMS.com to find out how to engage. You can expect much more innovation in this space from us; look out soon as we launch a new ‘Risk Labs’ page and showcase early innovations while continuing to deliver on the exciting roadmap of models and software.