One brain. All the hats.

The Chief of Staff analogy

A real Chief of Staff does not say “sorry, that’s Notion’s job” when you ask them what you agreed with Pierre last week. They do not say “that lives in Superhuman” when you ask what your inbox needs today. They do not say “I’d have to check Otter for the transcript” when you ask what was decided in Tuesday’s call. They just know. Because it is all in the same head.

That is what a personal AI has to be to be worth having. Not fifteen small clever tools sharing your data with fifteen different cloud companies. One system that carries your whole context and wears whichever hat the moment calls for.

Most people cannot afford to hire a Chief of Staff. The ones who can quickly find that a good one commands a six-figure salary, needs onboarding and holidays, and has to be trusted with the most private parts of your life. Ostler skips all of that. It never goes on leave. It never leaves the room. And every piece of context it holds sits on your Mac, encrypted, private to you.

The hats it wears

Every one of these is a category with its own subscription, its own login, its own privacy policy. Ostler covers all twelve from a single install on your Mac.

General AI chat

Ask it anything, but from a model that knows your history, your people, and your projects.

Instead ofChatGPT · Claude · Gemini · Perplexity

In-person conversations

Ostler captures the coffee, the dinner, the walking meeting. What you promised at brunch is remembered the same way as what you agreed in a Zoom call.

Instead ofPlaud · Rewind · Limitless · Bee

Digital meeting notes

Automatic capture and summary of your Zoom, Teams, and Meet calls. Action items and who-said-what, all transcribed on your Mac. No bot joining the call.

Instead ofGranola · Otter · Fireflies · Fathom

Personal CRM

Every person you know, deduplicated, enriched, and linked to every conversation you have ever had with them.

Instead ofClay · Dex · Monaru · Folk

Email triage

An assistant that reads your inbox, tells you what actually needs a reply, and drafts it in your voice.

Instead ofSuperhuman · Shortwave · SaneBox

Notes and personal wiki

A living wiki of your own life that compiles itself from every conversation you have, no manual tagging required.

Instead ofNotion · Obsidian · Roam · mem

Personal AI cloud

An assistant that knows your life and belongs to you, running on your own Mac instead of somebody else’s data centre.

Instead ofKin · Poke · Jo · Rewind

Research assistant

Ask, and it searches your own history first, the public web second, and cites both.

Instead ofPerplexity · Consensus · You.com

Writing help

Drafting, tightening, and tone-matching, from a model that already knows how you sound.

Instead ofGrammarly · Notion AI · Lex

Calendar awareness

Pre-meeting briefs, prep for the person you are about to see, and evening wrap-ups of the day.

Instead ofMotion · Reclaim · Cal.com AI

Tasks and follow-ups

Action items extracted automatically from what you say or write, not typed into yet another list app.

Instead ofTodoist · Things · Any.do

Reminders that matter

Birthdays, anniversaries, and the promises you made three months ago that you have already forgotten.

Instead ofNudge · Reminders apps · sticky notes

Stitched together: the PA you can finally afford

Triage, prep, drafting, follow-up, morning brief, evening wrap. Everything a private PA does, done by one system that already has your context. All twelve hats above, worn by the same head, so nothing gets lost between them.

Instead ofHuman PA · virtual assistant services · the twelve tools above

The meter is off

Here is the part cloud AI subscriptions do not quite tell you. The better the assistant is, the more you want to talk to it. The more you talk to it, the more it costs. Every good question is metered. Every long conversation is billed. Every large document you paste in costs money to think about. You develop a little internal accountant that pipes up every time you open the chat window: is this question worth two cents? Is this one worth three? And so you use it less than you would like to. You ration.

Enterprises are already exiting this trap. Jake Saper at Emergence Capital (early backers of Salesforce, Veeva and Zoom) calls it the tokenmaxxing hangover: cloud-AI bills quietly quintuple, and the company routes the work to cheaper local models when the invoice lands. Open-weight models already carry a quarter to half the volume across major aggregator platforms and are used by roughly 80% of startups. The excuse to run everything in the cloud is thinning fast.

Ostler does not have that accountant. Your Mac’s compute is already paid for – you bought the chip, and the chip does the work. The model runs on your chip. The answers happen on your chip. There is no per-token bill, no monthly reset, no context window you are paying to keep full, no premium tier that unlocks the good model. Throw everything at it. The cost does not move.

That is the difference between an assistant you use freely and one you ration.

The maths

A modern founder or operator who takes their work seriously is probably running most of this stack today. Not all of it, but seven or eight of them, quietly billed to a card every month.

Every one of those SaaS lines is a separate cloud company holding a slice of your data. A separate login. A separate privacy policy. A separate breach notification email waiting to arrive in a year or two. Consolidating them into one system on your own Mac is not just cheaper. It is saner.

The real magic: it all just works

Cheaper is nice. But the reason to consolidate is not price. It is the thing that happens when every hat sits on the same head.

Get Ostler – $99 Why you need this ›

Related reading

Ostler vs Granola · a longer look at one of these hats. Security · how twelve trust boundaries collapse to one.