Universal Input
One bar, wherever you add things. Text, voice, attachment, URL, slash command. AI Twin proposes where it belongs and waits for you to confirm.
Personal AI memory
Remember what matters. Find it when it matters.
AI Twin is a private, model-agnostic memory layer for your life. Upload what you have. Tell it what counts. Then ask for it when you need it, here or in an assistant you connect.
Start your Twin
- Rules and email
- Bring one thing
- Name it
Running a practice? Set up your own Twin first, then add the practice from the workspace switcher. It sits alongside, kept apart.
Everything is free for the first 7 days. No card needed. Add one when you decide to stay.
Not ready to start? See what your AI export holds. Free, on your device, nothing uploaded.
See what we capture, and what you controlPrivate by design. Audit logged. Export and delete, any time. Works with assistants that support MCP connectors.
How it works
What goes in. What it becomes. Where it stays.
Your life is scattered across files, messages, and inboxes. You choose what to bring. AI Twin types it, logs it, and stores it in the UK.
What you bring
One governed layer
People
Documents
Facts
Events
Episodes
- Row-level isolation, at the database
- Append-only audit log you can read
- Stored in the UK
- Your content is not used to train AI models. Our model providers' terms do not allow it.
- Captures wait for your confirmation
How you use it
An assistant that supports MCP connectors, such as Claude, or ChatGPT on a paid plan with developer mode on. Your memory stays in AI Twin. Model-agnostic. Export any time.
A connector, so your memory travels to the assistant instead of the other way round.
- Input
- Documents and PDFs
- Becomes
- Documents, and the Facts and Events inside them.
- Status
- Live. Contracts, letters, reports, manuals.
- Logged
- Every capture writes an audit entry you can read. Stored in the UK.
On-device demo
Watch a memory come back. Nothing you type here leaves your device.
Three sample memories. Ask a question in plain English. AI Twin ranks them on your device using an embedding model of about 37MB that your browser downloads once and usually keeps in its cache. Best on Wi-Fi.
The demo appears as you scroll near it; the model downloads when you first use it.
What AI Twin does
A typed memory layer, built carefully.
Review before anything is kept
Nothing is saved until you confirm. Duplicates and conflicts come to you to decide.
Typed Memory
Every memory entity carries a category: People, Documents, Facts, Events, Episodes, and further types beneath them. Queryable, editable, versioned, audited.
Hybrid Retrieval
Typed lookup for precise facts, vector search for meaning, full-text search, and facts that know when they were true, combined into a single answer.
Cards
Cards keep the records you look for most: policies, documents, things you own. Each one is live from memory, not a copy.
Workspaces
Personal Workspace and Practice Workspace, with clean scope switching. People you share with see only what you allow.
See it
Your memory, laid out.
Rooms
- Home
- Plans
- Things to keep
People
- Anjali
- Sam
- Rose
Timeline
- A conversation
- A decision
- A reminder
WHERE IT SITS
Swap the model. Keep the memory.
Many AI products keep what they learn about you inside their own product. AI Twin sits one layer above the models and holds the memory, so you can change the model without losing it.
You
AI Twin
Your memory. Typed, audited, yours. It stays here.
Your assistant
Claude, ChatGPT with developer mode, or another assistant that supports MCP connectors. Swap any time.
you bring it, once
Your apps and files
You bring what matters.
The model is a commodity. The memory is the point, and it is yours.
What we are not
Seven things AI Twin is not.
Defining a product is easier when you also say what it is not.
- Not a chat wrapper. You can chat with your Twin. But the product is the memory underneath: typed, audited, yours. Chat is one door in, not the whole product.
- Not always-on. AI Twin captures only what you ask it to capture.
- Not a memory model. The model is the LLM that writes the answer. The memory is yours, typed, held in our database. They are different things.
- Not vendor-locked. Your data, your rules. Export any time. Take your memory with you.
- Not a developer API. AI Twin is a product, not backend memory infrastructure we sell to other AI builders. Memory enters through native surfaces and travels out to the assistants you choose through the Model Context Protocol connector, under your control and audited.
- Not a plug-in. AI Twin is a destination product you sign in to, not a memory layer bolted onto someone else's chat window. Tools that help you bring memory in sit around the product; they are not the product. Your memory lives in AI Twin and leaves only when you choose to take it with you.
- Not loud. No exclamation marks. No urgency. No productivity theatre. Calm by design.
The context
Built for what professionals already need.
61%
of UK lawyers use generative AI in their work
LexisNexis, The AI culture clash, September 2025
17%
say AI is fully embedded in their firm's strategy and operations
LexisNexis, The AI culture clash, September 2025
AI Twin is built around audit, provenance and consent.
Built to UK GDPR and EU AI Act standards.
Lawful basis, data minimisation, storage limitation, audit, your rights.
Limited risk, as we understand it today.
The EU AI Act requires us to tell you when you are dealing with AI. Every chat reply and every Brief carries an AI label. We will not claim more than is true.
How AI Twin uses AIStored in the UK, processed in the UK and EU
Your memory and files are stored in the United Kingdom. AI inference, speech to text and email are processed in the European Union. Every category of provider we use, with its region and agreement status, is listed on our sub-processors page.
Use cases
Built for the bits of life that get lost.
Work
You uploaded the supplier contract two months ago. When the renewal email arrives, ask your Twin and it can bring back what you negotiated last time, so you can decide what to push on now.
Family
You captured the school’s term dates and the dietary letter from the nursery. When you ask “what’s on this week”, the answer can draw on them.
Health
You added your medication, your last blood test, and the GP’s notes during onboarding. When you prepare for the next appointment, the picture is already there.
Caring
You captured the care plan, the medication schedule, and the questions you wanted to ask the consultant. When the day comes, you walk in prepared, not piecing it together in the waiting room.
Founder life
You uploaded the investor deck draft, the cap table, and last quarter’s numbers. When you sit down to write the next update, your Twin has the story so far to draw on.
The stuff you saved
That article you bookmarked. The boiler manual. The warranty PDF. The brochure from the holiday you nearly booked. Your Twin keeps it findable, in context, when you actually need it.
Illustrative examples.
Why AI Twin is different
Same question. Different answer.
An assistant without your context
- You
- What questions should I ask the GP this week?
- Response
- Common questions to ask your GP include: how often should I have a checkup, what tests do I need, what lifestyle changes should I consider...
Doesn't know your medication, your history, or what changed since last time.
AI Twin
- You
- What questions should I ask the GP this week?
- Response
- Based on what you uploaded and told me during onboarding. You started a new medication six weeks ago. You flagged a side effect last week. Your last blood pressure reading was 145 over 90. Here are five questions worth asking, in order of priority.
What it knows about you comes only from what you brought to your Twin.
Illustrative example.
Trust
Your data. Your rules. Calibrated honestly.
AI Twin is built around your control.
- Your data is encrypted in transit and at rest. Access is logged and auditable.
- You can see what your Twin has stored, edit it, export it, and delete it. The export is a portable file you can take anywhere.
- If your trial ends, export and close stay open. Always.
- Deleting your account soft-deletes your data immediately and permanently anonymises it within 30 days. Operational backups roll off automatically.
- We do not sell your data. We do not share it with advertisers. Your content is not used to train AI models. Our model providers' terms do not allow it.
- AI Twin is model-agnostic. You choose the assistant you want to use. If you change your mind, your memory comes with you.
Built to UK GDPR and EU AI Act standards.
Pricing
Two ways to use it.
Start with Personal at £18 a month. Add people or a practice when you need to.
Personal
£18 a month
Your Twin, your memory. Includes one free additional slot, read-only, so you can bring one person in at no cost. Each further person is £9 a month, read-only or read/write, changeable at any time.
Practice
£43 a month
£18 Personal + £25 Practice. The owner is included; each person you add is £9.
A shared workspace for your practice. The owner is included at no extra cost. Each person you add is £9 a month, read-only or read/write, changeable at any time. Everyone shares the owner's Twin; anyone who wants a Twin of their own takes their own Personal subscription.
Everything is free for the first 7 days. No card needed. Add one when you decide to stay.
Prices are in pounds sterling. AI Twin Limited is not currently registered for VAT, so no VAT is charged. If we register for VAT later, the price you pay will not change.
Annual: two months free.
Questions, answers
What people ask, before they sign up.
Get started
Your Twin is ready when you are.
- Rules and email
- Bring one thing
- Name it
Running a practice? Set up your own Twin first, then add the practice from the workspace switcher. It sits alongside, kept apart.
Everything is free for the first 7 days. No card needed. Add one when you decide to stay.
Featured on
