Agent Cargo builds AI agents for shipping. When we joined, there was no brief and no product yet. The founder had deep domain knowledge and a clear conviction about the problem. Our job was to turn that into something users could try.
We worked with the founder to define a go-to-market MVP: not the full vision, but the smallest version that could get in front of users and generate real feedback.
We built user personas from conversations with people in the industry, mapped where their shipping workflows actually break down, and translated those pain points into user flows.
We designed a skeleton MVP first and iterated on structure and flows while changes were still cheap. Once the foundations held up, we refined it into a polished, demo-ready product.
We demoed the product to users, gathered usability feedback, and made improvements based on how people actually used it.
This is the kind of engagement where the work starts before the design does. Defining the product was part of the project, not a prerequisite for it.