Building AI Agents for Enterprise Operations

Happy Robot co-founders Pablo Palafox and Luis Parra met on their second day of college and have been building complex systems together ever since — including submarines for underwater robotics competitions. That appetite for hard problems led them to logistics, where Parra’s brother, a CFO at the world’s largest olive oil distributor, was hiring interns to call truck drivers because Walmart kept asking “where is my shipment?” The founders realized that the limiting factor wasn't willingness to buy — freight brokers told them repeatedly that a voice agent that could negotiate rates over the phone would be a no-brainer — but whether the technology could actually work. In late 2023, they decided it could.

Voice as the unlock, not the product

Voice was the wedge into supply chain, but Happy Robot quickly discovered that the real problem was not voice quality or latency. It was enterprise coordination. A customer asking “where is my shipment?” triggers a chain of actions that no single model can handle alone: checking the airline’s website, sending an email when no update appears, calling repeatedly when the SLA is at risk, and escalating across systems. The founders learned that raw model intelligence was not the bottleneck. What mattered was orchestrating work across fragmented systems, teams, and channels.