AI agents are coming to transform small business lending

Imagine a small business owner logs into their online banking portal to figure out their cash position -- a constant struggle for a large majority of them. A chat bot pops up, asks a few questions. After some responses, the business is approved for a working capital loan, and a few minutes later, the money shows up in their account. A dream come true for a small business. 

Now, compare that process to what really happens.

Small business owners have to go through a lot of paperwork, questions and answers for days -- if not weeks. And then wait! While at the same time, they have a business to run. The key here is that even after going through the process, there is no guarantee of approvals. The loan application may be denied or only partially approved, and then you might have to start the process again with another lender. Online lenders could be an option but may not be the most cost-effective. For a small business owner, the choice is between getting access to funds quickly at an excessive cost, versus the uncertainty of not getting funded at all. The answer is obvious. 

How can we get from current reality to the desired state?

Streamlining approvals and funding  

Lending is one of the key profit centers for a bank, but it’s highly regulated and scrutinized. Banks do want to lend money, but the bottleneck revolves around compliance and risk mitigation. This is where an agentic application can come in and make it easier for a lender to assess risk while protecting data privacy. AI agents can be deployed across the loan lifecycle, from underwriting to portfolio monitoring, thereby improving turnaround time and potentially instant funding. 

Underwriting Agents 

Underwriting is a tedious process for lenders and small businesses. In reaching the desired state of fast funding, AI agents can come to the rescue of underwriting teams by helping assess the viability and capacity risk of any small business irrespective of the industry type and geolocation. AI agents can streamline the entire application process, allowing lenders to say ‘yes’ more often. 

• Data Orchestration: Aggregating, correlating and analyzing data from multiple sources in real-time to uncover deeper insights into the business, including fraud risk. Whether it is tough to assess newly established businesses or those owned by people from underrepresented groups, such as minorities, everyone can be objectively assessed.  

• Personalized Needs Assessment: Agents can identify financial needs, based on projected capacity and the business growth needs, through data-driven questioning and customized lending flows for appropriate product, terms and interest rate offerings.

• What-if Scenario Planning: Assist underwriters in preparing financial ratios and cash flow projections, allowing them to quickly evaluate the impact on business growth and repayment behavior based on different economic scenarios.

Enabling elevated customer experiences

For existing bank customers, agents can completely transform the loan application process paradigm. Instead of a small business applying for a loan, agents can automate proactive funding based on business complexity, sales and market dynamics rather than a simplistic score cut-off and amount threshold approaches that exist today. 

Moreover, AI agents can act as a ‘banker on-demand’ and provide real-time industry and business specific insights based on the market conditions, enabling advice centric relationships. These agents can prove to be a cost-effective way for lenders to serve all customer types, irrespective of their current relationship value -- elevating customer experience to a whole new level. 

In summary, the proliferation of agents could result in significant efficiency gains that can enable access to financial services for more small business owners, especially in underbanked and underserved communities. An important consideration, though, is aligning AI strategies with evolving regulations and proactively addressing any potential implementation risks. Agentic applications should operate under rules governed by humans and mature over time through testing and performance evaluation. This will enable robust and compliant use of AI agents, while realizing the desired state of fast, efficient and proactive funding for small businesses.