The term “embedded finance” refers to the integration of financial services and products into the business processes and platforms of non-financial entities. For example, an e-commerce platform that allows its customers to open a bank account, make a payment, or apply for a loan without leaving the platform is using embedded finance.
But embedded finance is not just about enabling transactions. It is also about providing decision support and insights that can help businesses optimize their operations and achieve their goals. In this blog post, I will explore how embedded finance can become a full “decision support agent” for businesses, using generative AI models to leverage the data from both the enterprise (e.g., enterprise resource planning [ERP], treasury management system [TMS]) and the banking systems.
I want to prove that there is an advanced notion of the “embedded” proposition: What characterizes embedded finance is not only that financial applications are consumed “as-a-service” in a transparent way to the user. The key element is that the user can find support from a decision support “agent” embedded with the operations executed.
With the transactional part of embedded finance, there is a button on the ERP (or TMS) dashboard for users to click and open a new bank account, make a payment, or ask for a loan without having to move to a different environment. While still working within the enterprise environment, the decision support agent portion provides useful information and insights before the user clicks to execute a transaction.
Both the ERP/TMS and the banking systems hold a history of the user’s transactions that can give the user insights on what to do and what happens next. Generative AI models could give the user useful advice before executing the operation, learning from past experience what the bank is doing and what is the best option before execution. By generating that knowledge from existing transactional data and offering it as an embedded capability, “embedded” then becomes not only the transactional part but—most importantly—the knowledge part.
A likely use case: a business that sells parts for automotive equipment manufacturers uses an e-marketplace platform with embedded finance services. The engineering, supply chain, sales, and financial managers run a sales cash forecast from their ERP to determine whether the company should invest in a new welding robot machine to produce and sell a new line of products. This triggers the embedded decision agent that returns this message:
After analyzing your sales data and your bank account balance, here are some suggestions for you:
You have enough cash flow to invest in a new welding robot.
Based on satisfaction scores of the marketplace members, we recommend that you search across these three vendors:
You can access each vendor’s catalogs from here.
You can issue your RFP from here.
You have an outstanding loan of US$150,000 with an interest rate of 5%. We suggest that you pay it off as soon as possible, as you can save US$500 in interest payments. You can use our embedded banking feature to make a one-click payment from your account.
You can submit a loan request to these three banks based on your credit score profile:
To repay the loan installments, you can access the receivables finance program launched by
Based on past sales data, your customer base has purchased new products from you with a level of recurring sales over 75%. We propose that you offer them a discount of 10% on their next purchase. You can use our embedded banking feature to send them personalized coupons via email.
As you can see, this is an example of how embedded finance can become a decision companion for businesses, using generative AI models to provide actionable insights and recommendations based on the data from both the ERP and the banking system. This way, embedded finance can enhance the business operations and outcomes of non-banking entities, creating value for both parties.
Interested in learning more about my research on embedded finance and how it can benefit your business? Read my latest report, Embedded Finance for B2B Marketplaces: Onboarding New Client Bases or download the report summary here. You can also contact me here—I’m always happy to discuss your needs and offer you a tailored solution. In addition, I’ll be moderating a discussion on this topic with Citizens Bank’s Director of Product Strategy and Merchant Services, Kavita Kurella, and U.S. Bank’s Product Head of Fintech Strategy & Partnerships Bryan Schneider at Aite-Novarica Group’s Commercial and Small Business Banking Forum on October 5th. Register here to join us in New York.