Every enduring company begins with a single, stubborn frustration (ideally felt by the founders themselves). For Dedalus Labs, it was this: building and deploying MCP servers shouldn’t be this hard.

When the founders (Cathy Di and Windsor Nguyen) began experimenting with AI agents, they kept encountering the same bottleneck. Everyone they talked to (and most of the MCP subreddit) were spending hours fighting with AWS and GCP configs just to get something live.

Profound was inevitable. AI users are still very much in the garden, every piece of data presented to us because a benevolent machine tried in earnest to solve our problem, unsullied by advertising and ‘promoted’ results.

How do you satiate advertisers’ desire to influence the AI to recommend their offering without ruining this beautiful new paradigm? Profound answers this riddle by giving advertisers a way to influence the zeitgeist without polluting the town square.