Most Swiss businesses already use AI through cloud services — but every query sends your data to a server abroad. Octop offers a different path: a self-hosted AI assistant that runs entirely on your own computer. No per-seat subscription. No data leaves your office. Just honest AI you can trust.

What Octop actually is

Octop is a self-hosted AI assistant platform for households and small teams. One process serves the web dashboard, CLI, IM channels, and cron automation, all sharing one control-plane database under ~/.octop/ (SQLite by default, PostgreSQL optional). Each user runs a personal team of specialized agents and switches experts per task; experts and knowledge bases can be shared with others on the same install.

Give any agent one of 16 MBTI-style personalities through an interactive quiz. Beyond chat, Octop Memory gives agents hierarchical recall with full-text search, so memory travels with the workspace instead of staying tied to one chat. Chat reaches you through the Web Dashboard, Telegram, or any platform you connect.

Why it matters for a small Swiss business

Two things usually stop a small office from using AI on its documents: cost and data protection. Octop is free and open source, and it runs on a computer you already own — an office PC, a NAS or a small server. There is no per-seat cloud subscription to pay.

Under Swiss data protection law, the big question is always where your client data ends up. A tool that stays inside your own four walls is a far easier story to tell clients than one that sends their files to a server abroad. For doctors, notaries, fiduciaries and anyone handling sensitive documents, that is often the deciding point.

What you should know before you start

The honest catch: Octop is technical software. You install it with Docker — a standard way to run programs in a tidy container — or with a single-file “Lite” version. That is simple for someone comfortable with computers, but it is not a one-click app-store download.

Second, the “brain” is not included. You connect it to a language model — either a cloud one like OpenAI or Claude, which you pay for, or a local model you run yourself, which has no fees but needs a reasonably powerful machine.

  • Best for: teams with many recurring questions and shared knowledge bases.
  • Not for: a quick one-off question about two files — a direct upload is faster.

A sensible way to try it

Start small. Take one recurring task — for example, answering the same client questions week after week — and set up an Octop agent for it. Use it for a week and watch how often it saves you from typing the same answer again. That is the strongwinds way: no hype, no big bang. One real problem, one honest test, and a decision based on what you actually saw.