Meet the founders
Shoti is built by people who felt this problem before they built for it. Three founders, one loop: see what everyone’s building, build from anywhere.
Why we built it
Every role on a team runs agents now. The building got faster; seeing what everyone is building did not. We wrote down the whole argument — where it started, where the thinking turned, and where it arrived — in three essays.
If you would rather just talk: book twenty minutes with Abry, or write to hello@shoti.ai.
Questions
What is Shoti?
Shoti is the workspace where your team and its agents build your company — the people, the AI, and the product in one place. It draws the company as a city a person can hold in their head, lit where someone or someone's agent is working right now, and a narrator tells the story of the work when you return. Click a place and your agent starts there with the real files and context instead of a blank chat box.
What does 'humans in the middle' mean, and how is it different from 'human in the loop'?
Human in the loop describes a checkpoint added to a machine-to-machine pipeline: the system pauses where a designer decided it should, and a person approves so the pipeline can continue. Humans in the middle is a different shape — agent → people → agent. An agent shows what it is about to do, people decide with their own agents helping them, and the decision goes back at the pace the agents move. The person is not an exception handler inside someone else's pipeline; the person is where decisions live, and the layer exists to bring each decision to them while it is still cheap to make.
Is Shoti a multi-agent orchestration tool?
No. Orchestration frameworks coordinate sub-agents inside one person's process, which is the right tool for one person's task. Shoti is the layer between many people's agents, across tools and machines. An agent can execute a goal but cannot own one — direction, taste, and accountability live in people — so Shoti routes each agent's intent through the team's people and carries their decisions back, rather than letting agents coordinate the company among themselves.
Is Shoti a company brain or memory tool?
No. Brain tools hold what a company has captured — the docs, decisions, and product context: work that has left a record. Shoti is the layer beside them, carrying the state that has no record yet — the intent an agent just accepted and the files it is about to touch — while a teammate can still redirect the work. The roadmap includes brain integrations so context and decisions move both ways.
Does Shoti replace Cursor or Claude Code?
No. Shoti is the neutral layer above the coding tools, across every vendor and machine a team uses. You understand the state of the work from the city and its narrator, and the same compilation that told you the story becomes the context an agent starts from when you dispatch work back into your own tools. The catch, the signal that flags redundant work before a day goes into it, runs on the same record.
Why is Shoti a city map with a voice instead of a chat window?
Humans consume two channels natively: familiar scenes and language. Shoti renders a team's live work as a night city read at a glance (lit windows mean someone is working, two crews at one site mean trouble), and a narrator tells the story of the work on request, compiled only from the record of what agents and people actually did. Chat remains underneath, the way a terminal remains underneath a desktop; it is the raw layer, not the place a team lives.