The Andi Copilot Early Adopter Program: Guiding the Copilot
Welcome to Part Three of our ongoing in-depth look at the Andi Copilot Early Adopter (EA) program and its first major area of focus, the commercial credit and lending sector. Our blog series began with the origin story of Andi and an introduction to the EA program and its goals. Part two gave you a look at the steps we’re taking to ensure security and compliance.
In today’s post, we want to show you how we worked with our beta program participants to determine the best initial GenAI use case. That story also provides a compelling reason for why your institution should join the Andi Copilot EA program.
Andi Copilot EA Program guiding principles
GenAI is obviously an area of immense potential, but it’s one in which Q2 treads very carefully. The journey we’ve undertaken with the Andi Copilot EA program is one that continues to bring us closer to our customers and the problems they’re working to solve, while also deepening our understanding of how LLMs can add value to the financial industry.
As we develop this program, we keep the three guiding principles front and center.
-
- Our customers will have the loudest voice in determining which Andi Copilot use cases will rise to the top of the priority list.
- Our development process needs to be flexible so we can pivot to ensure we’re focusing on what matters most to the customers.
- We will maintain a healthy respect for the limitations and risks of this AI technology and take steps to mitigate those risks throughout the process.
The POC … and then the pivot
As noted in our first blog, our initial pass of research identified the commercial lending process as the first GenAI area to explore with Andi Copilot. Leveraging our industry expertise and broad range of customer inputs , we collected a range of pain points and challenges in the loan origination portion of the commercial lending cycle to build out the initial functionality in our proof of concept (POC).
Following the presentation of the Andi Copilot POC, we conducted deep-dive interviews with beta participants, to better understand each of the pain points their institutions encountered during the lending process. We then asked the beta participants to rank the uses cases in order of priority.
We wanted to make sure the customer problems guided the tech development, and not vice versa. And in this case, they guided us away from our initial direction.
Our original hypothesis was that the loan origination portion of the commercial lending cycle would be the area where Andi Copilot could make the greatest initial impact. But beta participants instead said their greatest bottleneck was occurring later in the process, during underwriting. Specifically, they wanted help with the laborious task of verifying accuracy between the credit memo and the loan agreement – a process that often requires 3-5 iterations and involves 3-5 employees each time.
It was clear that this point in the underwriting process was the place to start; the use case where Andi Copilot could have the shortest time to impact.
In addition to underwriting, we see additional future opportunities to apply the Andi Copilot accuracy and consistency skill – in other points in the lending cycle in use cases beyond lending, like account opening.
What’s next
The beta participant interviews and voting also uncovered other exploratory use cases in loan maintenance and loan compliance and governance analysis. Just as we did with the loan accuracy and consistency use case, we’ll follow the same steps: collect feedback from the beta participants in determining whether to proceed with the use cases, and if so, the best path to move forward on with each.
Meanwhile, we’re continuing to push forward using Andi Copilot to make the loan process faster and more accurate. We’ll get into the details of how we’re doing that in the next blog in this series.
Why join?
By now you can see that the Andi Copilot EA Program is very much a two-way street with our beta participants. Q2 conducts research and testing, shares that information with the beta participants, gathers their feedback, and then adjusts the process accordingly.
It’s an iterative process in which the customer has a great deal of influence in determining which pain points Andi Copilot will tackle next.
If you’re looking to improve the speed and accuracy of your lending process, or if you’re more generally looking for ways that your institution can leverage AI, we encourage you to join the Andi Copilot EA program. To make your voice heard and have a say in the shaping of Andi Copilot, join today!