Building smarter finance functions: AI adoption tips for SMEs

As artificial intelligence (AI) continues to reshape business operations, small and medium-sized enterprises (SME) are increasingly exploring how to integrate it into their finance functions. At the Manchester and Liverpool Finance Leaders Networking Events Mike Hollifield, director at Baker Tilly Digital, shared grounded, practical advice from his experience helping SMEs across sectors implement (artificial intelligence) AI safely and effectively.

Q: What are the most common starting points for SMEs looking to incorporate AI into finance and operations?

A: The most common entry points are automating repetitive, rules-based tasks, such as invoice processing, cash flow forecasting and expense categorisation. These tasks are time-consuming but follow predictable patterns, making them ideal for AI-powered automation.

We usually recommend starting with areas where structured data already exists and where manual effort is high. Even small improvements can lead to measurable time or cost savings. Integration is key and tools like UiPath and Boomi that work seamlessly with existing enterprise resource planning (ERP) or finance systems can make or break a project.

UiPath is a leading global provider of robotic process automation (RPA) software, that enables organizations to automate repetitive tasks that are usually performed by humans. Learn more about Baker Tilly and UiPath here.

Q: What is your perspective on risk and governance when adopting AI in financial processes?

A: AI brings some new risks, like protecting personal data, avoiding unfair bias and making sure its results can be understood and trusted. But the basic rules for managing any technology still apply: be open about how it works, make sure it can be reviewed and stay in control.

In finance especially, AI should be used to support human decisions, not to replace human judgment entirely.

We encourage clients to frame AI adoption within their existing risk management frameworks, make sure models are auditable, decisions traceable and that there's always a human in the loop where needed. Also, involving finance and information technology (IT) leadership early in the process ensures both compliance and practicality for internal and external regulations that AI may not consider.