ZaidiCommerce

ZaidiCommerce explores AI-assisted commerce and merchant connectivity in Kenya. The starting point is a coordination problem: businesses, product information, APIs, payment rails, and public digital services exist, but they do not automatically work together.

The broader interest is in making those pieces more useful and accessible through agents. The present implementation covers only part of that ambition.

What exists

The project has an agent-facing bridge for discovering merchants, searching observed product information, comparing offers, and calculating shopping plans. Its source includes basket planning and a comparison interface.

The comparison separates the request from the product identity. Two products can both satisfy a general request without being the same product. Package sizes and quantities still need to be calculated explicitly.

The boundary matters

Product records are observations. They are not a promise of current price, store stock, delivery availability, or a complete merchant catalogue.

The current project does not provide a live end-to-end purchase flow. Purchasing and payment work remains gated, and merchant checkouts remain separate. A useful shopping plan is one step in the process; it is not a completed transaction.

Still unresolved

Freshness, reliable availability, merchant-approved access, and the handoff from a recommendation to an order are separate problems. The project needs evidence for each of them before the archive can describe the broader ambition as an operational service.

This work connects to the business automation direction under ZaidiStudio. It is also a practical example of the relationship between software and everyday trade that runs through my interests.