What is Jootle

The shortest honest answer is this:

Jootle is your always-on AI, running on a private server we provision for you, holding only your data, reachable from anywhere you already work.

It is not a chatbot you visit. It is an assistant you live with. It remembers your projects and the people in them, it sits inside your email and calendar and chat tools, it works on things when you’re not watching, and it tells you what it did when you check back in.

If you’ve ever wanted “a Jarvis”, that’s the shape of it.

Who Jootle is for

Two kinds of people find Jootle useful.

Individuals who run multiple things at once and want an assistant that holds the whole picture. A consultant juggling four clients. A parent running a household plus a side business. A creator producing across podcast, newsletter, and YouTube. The common thread is that no single SaaS app is the answer, because the work spans email, calendar, documents, lists, projects, contacts, and ideas all at once.

Small teams and businesses who want a shared assistant for the operations of the business. Each team member can talk to it, it learns the business over time, and it does the kind of work that usually falls into “we should hire someone to do this” but never quite gets done.

Jootle is not the right fit if what you want is a single-purpose tool (“just a calendar scheduler”, “just a CRM”). There are excellent point tools for those. Jootle is for the work that lives between them.

What an instance gives you

When you spin up a Jootle instance you get, on day one:

A private server, provisioned for you. We spin up a dedicated VPS for your instance; your data lives in that server’s database, not in a multi-tenant system shared with other customers. Single-tenant by design.
A web app at your instance’s URL, with chat, projects, lists, ideas, artifacts, contacts, and goals already wired up.
Three default channels. You can talk to your AI through the web app, Telegram, and email. Add more later as you go.
A starter set of toolkits. Ideas, Lists, Forge, Finance, CRM, Game Arcade. You install more from the library when you need them.
Your AI, named. You give it a name on day one. Most customers pick something personal. The handbook generally calls it “your AI” or “Jarvis” interchangeably.

The mental model

Three things, woven together.

1. An assistant that knows things. Your AI builds a picture of your life: who Lily is and how she’s related to you, that A Box of Chaos is Jason’s company, that the kitchen renovation is a project with a contractor named Mike. This isn’t manual data entry; it’s something your AI accumulates from the conversations it has with you. The picture gets richer over time, and your assistant uses it to answer questions you didn’t bother to spell out.

2. An assistant that does things. Your AI can act, not just answer. It can send emails, run multi-step processes, create documents, watch your inbox for a specific pattern, generate weekly reviews. Some of the work happens immediately when you ask; some of it runs on a schedule and reports back later.

3. An assistant in the channels you already use. You shouldn’t have to remember to “open Jootle” any more than you remember to open your contacts app. You text it in Telegram. You CC it on an email. You ping it from web chat while you’re working on something else. The same assistant, the same memory, behind every channel.

That’s it. Everything else in the handbook is “how each of those pieces actually works.”

What it isn’t

It helps to be explicit about the shape of the product.

Not a meeting scheduler. It can help with scheduling because it has your calendar, but that’s a feature, not the point.
Not a productivity app you tend to. You should not need to “use Jootle” the way you use Notion. You ask, it does, and you get on with the rest of your day.
Not a multi-tenant SaaS. Your data and your AI run on a dedicated VPS we provision and operate for you, with no other customer’s instance on the same server. We maintain the infrastructure; we don’t read what’s on it.
Not magic, and not “AGI”. Your AI gets things wrong sometimes. It will tell you when it isn’t sure. There is a whole chapter, “When things go wrong”, about how to correct it. The right mental model is “a competent new hire with a great memory and infinite patience”, not “an oracle that always knows”.

Where to go next

If you want to get going right now, read Your first 15 minutes. It assumes your instance is up and walks you through the small set of moves that make everything else easier.

If you want the one concept that unlocks the rest of the handbook, read Playbooks and goals. It’s the mental model behind every automated thing in your instance, and customers who skip it tend to phrase requests in ways their AI can’t quite act on.

If you’d rather wander, the full chapter list is in the outline. Most chapters are short and stand on their own.