How banks are using AI to lend to SMEs faster and more reliably

Automated lending with AI and Open Banking

Two technological innovations enable Banca AideXa to overcome the business impediments of SME lending: Simplified data acquisition through Open Banking and powerful data processing using AI analytics.

Open Banking, in this context, refers to the gathering and processing of the borrower’s account information via standardized programming interfaces (APIs). This allows the processing of diverse transaction information, providing a comprehensive picture of a borrower’s financial situation.

For instance, Banca AideXa’s clients agree to provide access to their current transaction history of business accounts. „Instead of relying solely on figures from financial reports, we can utilize thousands of data points as a basis for decision-making on each credit application. This bypasses the bottleneck traditional financiers face in SME lending,” explains Matteo Camelia, Head of Data Science at Banca AideXa.

This data is too extensive to evaluate using either credit analysts or traditional statistical models within an acceptable timeframe. Hence, the bank employs various machine learning analysis models to create detailed profiles of their clients. This enables conclusions about a client’s solvency that go beyond the information of financial statements.

Within 20 minutes, applicants receive a response regarding the feasibility of their lending request. Upon approval, the bank transfers the funds within 48 hours. The speed does not compromise risk; the decision-making quality of the analysis models is high, with an unusually low default rate. Only 3.4 percent of all payments are non-performing, a remarkably low figure compared to financial institutions deciding on credit disbursement traditionally, emphasizes Matteo Camelia.

Why AI Complements Rather Than Replaces Human Expertise

Does the use of AI render credit analysts obsolete? Not at all, assures Matteo Camelia: „In over 95 percent of cases, the decision is made automatically. However, for larger credit applications, it has been shown that the interaction between software and humans is even more effective than a purely machine-based approach.”

If a final decision cannot be made automatically, a specialist reviews the application additionally. The decision-making platform provides the necessary information to the credit analyst in a structured format via a specialized user interface, with particular emphasis on the traceability of the machine’s decision. „The models are not a black box. We can always trace the decisions if needed and explain to clients why a rejection occurred,” says Camelia. „Humans can still relate to humans better. The tandem of human and machine is the key to success.”

The Right Technology for AI-powered SME Lending

While Banca AideXa’s self-developed AI analysis models are crucial for the fintech’s individual success, a central component in the technological fundament of the digital decision-making process is the ACTICO Credit Decision Platform. It facilitates the implementation of the bank-internal Policy Rules, and various scoring models and orchestrates the data gathering for rule execution.

The data entered by customers in the frontend, as well as externally sourced data, are seamlessly transferred to the Credit Decision Platform via customizable APIs. This platform processes the data for utilization, evaluates them, and makes decisions based on Banca AideXa’s rule models. If an automated decision is not possible, the application is handed over to an analyst. The platform features an underwriting user interface, allowing credit analysts to review the case.

The input data, as well as the generated results, are provided to the CRM system and to the Banca AideXa data lake. They are available for customer service and for the continuous development of analytical models.

„ACTICO plays a central role in our technological infrastructure, as the platform can be implemented very flexibly. The implementation is the crucial step, for without it, Open Banking and AI-supported models would be nothing more than an academic exercise. Thanks to ACTICO, it became the digital workflow as we use it now,” says Matteo Camelia, Head of Data Science at Banca AideXa.

Access to platform technology was as important as its rapid, productive deployment. „For us as a startup, a short time-to-market is crucial so that we do not unnecessarily spend startup capital. ACTICO Professional Services precisely enabled that. They supported us throughout the entire development phase with all their expertise, enabling us to launch our automated lending very early in regular business operations,” concludes Matteo Camelia.

Conclusion

Artificial Intelligence and Open Banking are paving the way for automated lending as the new standard in SME financing. With the right platform, financial service providers can automate the assessment of loan applications through a fully digital decision-making process within just a few minutes. The success story of Banca AideXa serves not only as a beacon for other fintechs but also for traditional lenders.

Software platforms grant them access to extensive data sources on applicants through open APIs. Automated decision-making routines evaluate the applicant’s creditworthiness in accordance with internal policies nearly in real-time, increasingly with the support of AI, yet, always with the option for a scrutinizing review by human credit analysts. This enhances transparency, speeds up the procedures, and reduces cost.

The ACTICO Credit Decision Platform provides all the necessary components for this and seamlessly integrates into existing IT landscapes. Through precise implementation, lenders can cost-effectively modernize the previous financing of SME customers and transform it from a side business into a promising growth venture.