TrueNorth Intell

Your data, your system, your call.

You are about to let something into the working parts of your business — the email, the calendar, the books, the customer list. That deserves a straight answer about where it goes and who can see it, written in language you can actually check.

This is that answer. If anything here is unclear, ask me directly and I will tell you.

The short version

Your system runs on hardware you own, under accounts in your name — and anything I hold for you, I move into your name when you ask. You own it outright — if we part ways it keeps working and you keep it. Nothing goes out to a customer, a vendor, or the IRS without a person approving it first. Your business data is never pooled with another client's, never sold, and opted out of AI model training on every account it runs through.

Where your system actually lives

On a machine you own

The system runs on a computer you own — your own desktop, or a server you own and pay for — using accounts in your business's name, including your own AI account, which you hold and pay for directly. Your business context and records live in ordinary files in a folder you control.

I keep my own working notes about the build, in my own private systems. The standing access I use for updates and repairs — a login key on your server, a seat on your code repository — I put in writing for you, along with how you end each one. Anything else, you let me into for that session.

Who else can see it

An AI system needs an AI provider, the way your books need a bank. Here is every outside company involved, what it sees, and what it has committed to in writing. The links go to their own terms.

Who What they see Their commitment
Anthropic
The AI doing the reasoning
The text of what the system is working on at that moment — the email being drafted, the job note being read. On a Pro or Max plan, chats are used to improve its models only if the account allows it, and that setting is switched off as part of setup, with narrow exceptions such as conversations flagged for a safety review. Business and API accounts are not used for training by default. Pro and Max · Business and API
Google
The AI that writes your daily brief
The business data the brief is built from — the numbers and notes it pulls together each morning. On the paid API tier your system is set up with, Google does not use prompts or responses to improve its products. Their terms
OpenAI
Only if you talk to your system by voice
The voice note, to turn it into text. Data sent to its API is not used to train or improve its models unless the account opts in. Their policy
Telegram
Only if your brief or messages come through Telegram
The messages between you and your assistant. Stored on Telegram's servers like any Telegram chat; bot conversations are not end-to-end encrypted. Their policy
GitHub
Only if your system updates from a private code repository in your name
The system's files and the updates I publish to it. A private repository under your own GitHub account and its terms. Their privacy statement
Your server host
Only if your system runs on a server you rent in your own name
The machine your system runs on. Governed by your own account with them.
Your existing software
Books, email, calendar
Only what it already sees — the system reads from the accounts you already have. Governed by the terms you already agreed to with them. Nothing new is introduced.
Your own AI account
In your business's name
Everything your system processes. Opted out of model training as part of setup. You hold the account and can revoke it at any time without asking me.

Those are the companies your system runs through. My own working tools — the recorder I use for our conversations, with your permission, my email, and the AI account I keep working notes in — handle what you share with me, on the clocks in the retention policy. Your data is not pooled with other clients' to build a bigger product, it is not sold, and it is not shared with anyone not directly involved in running your system. Each client's setup is kept separate from every other client's.

What it will never do

These are limits built into how the system works, not promises about intentions.

If we part ways

You own your system outright. That is the point of the whole arrangement, and it is written into the agreement rather than left to goodwill.

If you stop working with me, the system does not switch off or expire. It keeps running on your machine, under your accounts. Anything I hold for you — a key, a bot, a sign-in registration — I replace with one in your name when you ask, and I end the access I had. There is no platform to be locked out of, and no leverage held over you to keep you paying.

How long I keep anything

Short answer: as little as the work allows, for as short a time as the work allows — and on a client-owned system, most of it never exists on my machines to begin with. The specific clocks (what deletes when, what survives, and why) are published in full in the data retention policy. That page is the one our service agreements point to.

What I am not

I am one person who builds systems that live in your business, not a software company holding thousands of businesses' data on shared servers. That difference is why this page reads the way it does.

It also means I am honest about the edges. I do not currently take on work involving protected health information, and regulated-data engagements get scoped specifically rather than assumed. If your business has an obligation I would need to meet — a professional confidentiality rule, an industry requirement, a client contract of your own — tell me early and I will tell you plainly whether I can meet it or not.

Questions, or something here you want in writing? Email logan@truenorthintell.com. The full terms live in your service agreement, and I will walk you through any clause in it before you sign anything.