Kenyan Startup Fastagger Develops Offline AI for African Small Businesses on Mobile Money Platforms

Fastagger, a Kenyan startup co-founded by Mutembei Kariuki, Jude Mwenda, and Stephanie Njerenga, is pioneering AI solutions specifically designed for the technological realities of developing economies. Recognizing the challenges of slow internet, limited computing resources, and low-memory mobile phones prevalent in Africa, Fastagger developed AI models that can operate offline and on basic devices.

Their flagship product, Auni, launched in 2024, is a business-intelligence assistant integrated with M-Pesa, Kenya's largest mobile money service. Auni helps small and medium-sized businesses, such as salons and restaurants, analyze their transaction data to identify customer trends, optimize operations, and personalize marketing efforts. This innovation addresses a critical need for businesses that previously relied on manual record-keeping and lacked the tools for data-driven decision-making.

Fastagger's approach involved adapting open-source optical character recognition (OCR) models to accurately read financial statements from mobile money apps, overcoming issues with low-quality PDFs and unfamiliar formatting. Crucially, they ensured that data organization and basic analysis functions could run on-device without an internet connection, catering to users with unreliable connectivity. More complex queries, however, still require online access.

The startup's focus on mobile money data is strategic, given that over 80% of Kenyan adults use these services, providing a rich source of information. This localized approach to AI development, building solutions around existing infrastructure and user behavior, positions Fastagger as a key player in making AI accessible and practical for African small businesses, moving beyond adaptations of products designed for wealthier markets.