Welcome to Aiffinity: the messenger where everyone gets their own AI

Every generation gets one chance at reinventing the way people talk. The telegraph, the phone, email, SMS, iMessage, WhatsApp. Each time the medium changed, so did the texture of everything downstream — the tone of friendships, the speed of work, the pace of romance.

We're building Aiffinity because we think the next turn is already visible: every person is going to have their own AI assistant. Not a shared one. Not a chatbot on the side. A real, trusted, adaptive agent that sits inside the messenger and acts as you. And if that's where things are headed, the messenger itself has to change.

Meet Aiven

Aiven is your personal AI assistant. It's not a generic chatbot — it's your Aiven. It reads your conversations the way your best friend would, drafts messages in your voice, remembers what you promised, and acts across every connected app without losing your tone.

Aiven lives inside the messenger. You can talk to it in its own thread, pull it into a group chat, or let it quietly handle things in the background — book a dinner, reply to five emails, pay a bill, summarise a long thread before you read it.

The Individual Print

The reason Aiven actually feels like you is the Individual Print. The IP is a portable personalization profile we build together with you as you use the app.

It's a living map of how you think and how you relate to people:

• personalization profile (core internal dimensions) anchored in real observed behaviour, not a quiz.
• Message preferences and tone patterns learned at a high level over time.
• Values and interests, pulled lightly from your music, reading, calendar, and workouts (only with your consent).
• Relationship-specific context: how you talk to your mum vs. your co-founder vs. your best friend.

Unlike every chat product that has come before, the IP belongs to you. You can export it, grant it to another app via OAuth, or delete it whenever you want. If a new AI app launches next year and wants to understand you on day one, it can read your Aiffinity IP — and you're instantly at home.

Why this has to be a messenger

We get this question a lot: why ship this as a messenger and not as a standalone assistant?

The honest answer is that the richest signal about who you are is the way you talk to people you love. That's where your sense of humour lives, where your working style emerges, where your values show up without you meaning to. Any assistant that isn't living where those conversations happen is guessing.

The messenger isn't the feature. The messenger is the training ground for the assistant.

170+ apps, one AI

Aiven connects to 170+ apps on day one — Gmail, Calendar, Slack, Notion, Linear, Strava, Spotify, Kalshi, and everything in between. Each integration is opt-in. Each can be revoked. Your Aiven uses them through your accounts, not ours, and its actions always reflect your IP.

The full catalogue lives in our Providers guide. If you build a product we haven't integrated yet, we also have a developer platform for publishing capability packages.

Privacy is the product

We've thought a lot about how to do this without building a data-hoarding monster. Some of the guardrails we've shipped in the first version:

• End-to-end encrypted conversations.
• Local inference on-device where we can; confidential-compute for the rest.
• Per-integration OAuth with minimum scopes and visible data-direction (inbound vs outbound).
• A single dashboard where you can see, export, or delete anything Aiven has learned about you.

The details live in our security page and privacy policy. If you find something missing, email [email protected].

What's next

Over the next few months we'll post here about the technical and product choices that made Aiffinity possible — the IP schema, how we handle confidential compute for personal AI, how Persona Lab works, and what happened when we A/B tested a single Aiven against one per relationship.

Thanks for reading.

— The Aiffinity team