Pineapple handbook: running the company in the open
This handbook shows how we run Pineapple: who we serve, how we choose work, what the numbers say and where AI helps. It includes one real FAQ draft you can inspect and a workflow you can try in your own business.
Conceptual illustration. It does not show the Pineapple interface or a measured result.
New to Pineapple? Read what Pineapple does. Giving an assistant context? Use the AI starting point.
Want to try it? Start with the FAQ drafting exercise. You will finish with one short answer and a list of the facts supporting it. You can use the supplied Pineapple example before bringing your own business information.
That means giving AI a clear customer, a goal, useful evidence, and permission to do specific jobs. It also means keeping a record of what it tried, what happened, and where a person stepped in.
Our ambition is to give AI more responsibility as the evidence supports it. The pages distinguish current work from that ambition.
What exists today
Pineapple has a private operating repository called Peter Assistant. It contains our customer definition, business plan, reports, and recurring review workflows. AI helps inspect evidence and prepare work. Peter remains responsible for business decisions and consequential actions.
We have not proved that AI can run or grow Pineapple on its own. Full autonomy is an ambition. The useful test is whether customers get better outcomes while the business needs less avoidable human work.
This edition includes a dated billing scorecard and one real, checked FAQ draft. The worked example records the source conflict we found, the wording we prepared, and what remains unproved. The evidence is dated; this handbook is not a live dashboard.
Choose your starting point
If you want to… | Start here |
|---|---|
Understand the product, business model and customer journey | |
Give an assistant reliable business context | |
Decide whether Pineapple fits your business | |
Understand how we choose work | |
See where humans remain responsible | |
Understand the numbers and their limits | |
See how we approach growth | |
Inspect one actual result | |
Follow the general process | |
Try this in your own business | |
Understand what we share | |
Read the first progress note |
What we intend to share
Our customer focus, priorities, metric definitions, useful operating instructions, decisions, mistakes, and the limits of AI's authority. Results should include the period measured, what the evidence covers, and what remains unknown.
Customer messages, account details, credentials, and private business records stay private. A useful public explanation should not require exposing the people who trusted us with their data.
Try one task
Each operating page ends with a small action or a reusable example. Start with one workflow, run it manually, and compare the result with the work you do today. You do not need Pineapple, a developer, or an autonomous agent to use the method.
If you are here to build a website, the Pineapple product documentation covers the product itself. This guide covers how we run the company and what we are learning.
Updated 1 November 2026.
