The Ateneo de Manila University’s Business Insights Laboratory for Development (BUILD) is pioneering an initiative to integrate artificial intelligence (AI) as an augmentative tool for small business operations across the Philippines. Recognizing the dominance of small enterprises in the national economy, BUILD researchers have focused on improving how these businesses manage their everyday transactions.

The team, including Zachary Matthew Alabastro, Joseph Benjamin Ilagan, Lois Abigail To, and Jose Ramon Ilagan, concentrated on the widespread use of traditional pen-and-paper logbooks. These physical ledgers are a low-cost, reliable method for recording daily sales and inventory in settings ranging from food stalls in shopping centers to neighborhood sari-sari stores. Despite their dependability, these handwritten logs pose challenges when trying to analyze data efficiently.

Conversely, AI excels at processing and extracting meaningful insights from business data, such as identifying sales trends, monitoring product performance, and recommending inventory adjustments. However, many small business owners remain hesitant to adopt fully digital solutions due to concerns about steep learning curves and potential job losses.

To address this, the researchers advocate for a "copilot" model whereby AI supports and enhances human decision-making rather than replacing it. At the Artificial Intelligence in Human-Computer Interaction Conference 2025 held in Sweden, they presented their study showcasing how optical character recognition (OCR) combined with large language models (LLMs) can transform handwritten sales records into structured digital data.

Developed and tested in a food stall within Ateneo’s Student Enterprise Center, the system utilizes Python programming alongside AWS for OCR and Anthropic’s Claude 3 Haiku LLM for interpreting the handwritten entries. This approach enables even users without digital expertise to easily understand their inventory performance and sales trends.

By capturing photos of logbooks, the AI reads product names, matches prices, and generates detailed sales summaries. This facilitates quick identification of best-selling items or slow-moving stock, empowering small businesses to make informed decisions about restocking and pricing.

Although the prototype currently demonstrates moderate accuracy, the team is optimistic about future enhancements. The technology also promises adaptability, with potential uses extending to inventory management, delivery tracking, and payroll documentation.

Ultimately, the AI-driven tool aims to mirror the pen-and-paper logbook’s simplicity and affordability while gradually providing small businesses access to insights traditionally available only to larger companies. As AI systems improve in understanding local handwriting styles, Filipino micro-entrepreneurs stand to benefit significantly from this accessible innovation.