How I Got Here, and Where Retail Is Going
Ever since working on Paystreak.co (scan & go self-checkout for our stores) years ago, I had been thinking a lot about how customers will interact with retail (in my case c-stores) in the future. Back then, in 2022 and 2023, AI was just rolling out as a chatbot, and I wasn’t incorporating it in anything. Still, I knew that how people were interacting with retail wasn’t going to stay the same for very long. Everything from finding products to checking out was going to change. Given the rapid advancements AI has made since then, it’s become even more clear that we’re entering a new paradigm. It’s not here yet, but it’s about to be.
I first began thinking about and building StreetTea with a focus on gas prices. How would autonomous vehicles really know what gas station or charging station to go to? Yes, Google Maps, but how would it optimize for its occupant? Products, pricing, promos, specific rewards, brands, favorite locations, etc. These customers are also convenience store customers, and if these systems will be optimizing for such, it will need to know what is actually in the store.
Last year, I was working with a distribution company on a system that could help them understand what is in their customers stores and how they could better assist them with planogramming, merchandising, and sales, all via photo. They ended up scrapping the project, but all that experience directly applied to the problem I was looking to solve with StreetTea. With a whole year worth of improvements in AI, the problem was solved with the same mechanism: reading what’s in a store straight from photos of its shelves.
Right now, StreetTea is a platform that develops machine-readable digital storefronts for retailers. It functions as a local SEO and discovery tool that is tailored specifically for AI search engines and AI agents or “agentic discovery”, but the ultimate goal is far more. I want StreetTea to solve for how agents will interact with, find, and ultimately transact with retailers. Whether that agent is an autonomous vehicle, your favorite AI chatbot, or one of many personalized agents that will no doubt come to market. Instead of fumbling with a dopey app to scroll and add to a cart, this process will shift to natural language commands, and I aim to create the layer that facilitates such a request.