How Building Markets is Leveraging AI for Small and Medium Enterprises

How can Small and Medium Enterprises (SMEs) derive real value from AI?

Elizabeth Brown, Executive Director of Building Markets, shared insights in the June 2025 Applied AI (XD131) course, including how nonprofit like Building Markets is thoughtfully integrating AI to scale its impactful work, while emphasizing that trust and verified data are the foundation for any successful AI solution — especially when connecting real businesses to real opportunities.

Slide from Building Markets presentation

About Building Markets

Building Markets is a 20-year-old global network that supports SMEs through data collection, training programs, and direct market access. Operating in regions like Colombia, Latin America, the Middle East, Türkiye, and Southeast Asia, their core mission is to empower SMEs, which constitute 90% of firms globally and are vital for sustainable development, by helping them overcome barriers to market access, especially in procurement and export.

Their unique "find, build, connect, advance" model guides their operations:

  • Find: Collecting relevant data on small businesses.

  • Build: Providing tailored training programs focused on procurement, export access, digital presence, and financial management.

  • Connect: Directly linking buyers and suppliers using the collected data.

  • Advance: Supporting the overall growth of SMEs.

Since 2004, Building Markets has achieved significant impact, verifying over 28,000 businesses and facilitating $1.5 billion in transactions, leading to the creation of 75,000 jobs. Their work specifically targets economically vulnerable communities, including refugees, women, and youth, by providing them with capacity building and market access.

Slide from Building Markets presentation

The AI Transformation: Intelligent Matchmaking

Historically, buyer-supplier matchmaking at Building Markets was a manual and time-consuming process. Teams on the ground would manually search their system to connect buyers with relevant SME suppliers. To address this and scale their impact cost-effectively, Building Markets has partnered with Exchange Design to develop an intelligent matchmaking service.

This is a chat GPT-style tool that searches Building Markets' verified database to surface high-quality SME suppliers that meet a buyer's needs. The aim is to significantly reduce costs and time for their teams while simultaneously increasing the potential for more matches and greater impact.

Looking ahead, Building Markets envisions expanding their AI applications to include:

  • Automating inbound lead generation for SME data through APIs and partnerships.

  • Proactively suggesting business matches based on buyer and supplier profiles.

  • Expanding through an affiliate model by partnering with local organizations worldwide to rapidly grow the SME data within the tool. This scale model aims to create a comprehensive platform for global SME matchmaking, ensuring trust and verification of data by local partners.

Slide from Building Markets presentation

Key Takeaways from the Discussion: Data Quality and Iterative Development

Following the presentation, participants asked questions and discussed their own experiences, producing the following key insights:

  1. The Primacy of Data Quality and Verification: Inaccurate information in AI tools can erode trust, even with legitimate connections. Verification of vendors is a unique and paramount aspect of Building Markets' system, ensuring that businesses are real, formal, and trustworthy. This quality of data differentiates their tool from general AI chatbots.

  2. Lessons Learned in Chatbot Development: Consider shipping a functional prototype quickly (even at 80% completion) and then iterating based on user feedback. Consider co-creating the tool with the people who plan to use it with a focus on transparency and adaptability, recognizing that the technology is moving quickly and partners need to understand how the AI works to solve their needs.

This series features insights from our June 2025 Applied AI (XD131) course. Listen to the AI-generated podcast for an audio recap of this event.

Next
Next

Rethinking Openness in the Age of Foundation Models