Hivemind started with a bet that sounds obviously right: retail investors lose because they lack the data professionals have, so give them a terminal. The same supply chains, the same filings, the same news, at a price an individual can pay.
We spent about a year before going full time experimenting with financial datasets, news streams, graph retrieval and entity resolution, and then three months building wide. Screens for companies, screens for sectors, screens for funds. That is what a terminal is supposed to be.
Thirty conversations
In April 2026 we spoke with about thirty people: retail investors, hedge fund managers, investment professionals and bankers. The retail investors told us where they lost time and confidence. The professionals told us which information actually moves a decision and which is noise.
The terminal idea did not survive. Retail investors are not starved of data. They are drowning in it. Someone already running three apps, a newsletter and a Discord does not need one more screen, even a cheaper one.
What people asked for was much smaller. What happened to my portfolio? Why did it happen? What should I be watching? Three questions, first thing in the morning.
The second thing we heard: people do not pay analysts for data, they pay them for direction. Every retail tool stays away from that, because direction starts to look like advice. We think the honest answer is to lay out the evidence so clearly that you can reach the conclusion yourself: rank what matters by how directly it reaches you, show both ends of a supply chain, and back every claim with its source.
What changed, and what did not
A terminal makes sense for someone whose job is watching markets. Most people with money in the market are not that person. They open an app for a couple of minutes because something moved. So Iris shows one thing at a time: an event, the chain that connects it, the holding it reaches.
The engine did not change. The knowledge graph of suppliers, customers and competitors that we built to power the terminal is the same graph that now explains why one of your stocks moved. The surface changed, not the engine, which is why the change took a week instead of a restart.
The expensive lesson
The most expensive thing we did was build for breadth. We spent three months on wide coverage because that is what a terminal has, and the people we built it for touched one screen. We had built for an investor we imagined. Iris is built for the one we met: someone with a job, a portfolio and a few minutes in the morning.
If that sounds like you, the Daily Brief is where to start.