Everything you need to know about Shogo

Shogo is a self-evolving AI agent. You describe the app, workflow, or system your team needs in plain English, and the agent builds it — the screens, data models, automations, and actions — live in minutes. The same agent then lives inside what it built and keeps improving it as your needs change.

A chatbot answers questions and then forgets. A workflow builder makes you map every if/then step by hand, and it breaks when your data shifts. Shogo is goal-based: you give it the outcome you want and it reasons out the path, builds real software you own, and rewrites that software when you ask — no rigid node graphs, no starting over each chat.

Shogo generates real, type-checked code. Every model you describe becomes a full set of type-safe API routes and a typed client — generated, not guessed. Each tool call the agent makes is validated against a schema before it runs, and the build loop fails fast so the agent fixes an error before it ever ships to you.

Yes. Shogo keeps persistent, per-user memory indexed with SQLite full-text search — retrieval runs in single-digit milliseconds with no separate vector database. Instead of forgetting the moment a task ends, the agent carries context forward across the apps and systems it builds for you.

No. You build and change apps with natural-language instructions — describe what should exist and the agent assembles it. Because Shogo produces real, ownable code underneath, developers can dig in when they want to, but no one has to.

Tell the agent what should change — a new field, a new view, a new workflow, a new report — and it rewrites the app while you keep working. No new SaaS tool, no consultant, no admin console. The apps improve themselves as you ask.

Shogo routes across models through the Hoshi model router: one key reaches Anthropic, OpenAI, Google, and local models. A sub-agent router picks the cheapest capable model for each task in under a millisecond, so routine work stays cheap and hard problems get frontier-grade reasoning.

Yes. Shogo has 1000+ integration bridges into the tools, data, and workflows your team already runs on. It doesn't just move data between them — it reasons over that data and takes informed actions inside the app it built.

There's a free tier to start, then paid plans are transparent and per-seat: Basic at $8/month (single user), Pro at $20/seat/month, and Business at $40/seat/month. Paid plans are unlimited usage within fair-use rolling windows — no credit pool to drain.

Yes — that's the core difference from SaaS. You own the apps, workflows, and data the agents create. Cancel anytime and, unlike a subscription tool, the software your agents built doesn't disappear. It's yours.

Paid plans are unlimited within rolling 5-hour and weekly fair-use windows. If you choose to enable overage, anything beyond your window is metered at the raw provider cost plus a flat 20% markup, up to a hard cap you set. You can disable it at any time.

Legacy RPA platforms carry seat, bot, and orchestrator fees that push real first-year cost into the tens of thousands, plus months of professional services. Shogo starts at $8–$40 per seat with software you own — and your first app is live in minutes, not quarters.

Business plans add team-grade controls: SSO, audit logs, per-member spending limits, and usage analytics. Because Shogo produces real code and infrastructure you own, you're never locked into an opaque black box — you can see and control what runs.

Yes. Shogo's runtime and SDK are open source, and you can pin a project to a paired desktop or self-hosted worker. Self-host the API, the gateway, the database, and the storage — the same agent and artifacts run on your VPS, your desktop, or Shogo Cloud.

They stay yours. The code, the data, and the infrastructure the agents built don't vanish when a subscription ends. No SaaS lock-in and no vendor roadmap holding your operations hostage.