AI works well when it has the right context, but often that context is spread across platforms without an easy way to access it.
DataConnect changes this: it is an open-source desktop utility that pulls cross-platform data in one click. Download the app, connect your accounts, and DataConnect handles the rest: running on your machine, storing everything locally by default.
It currently supports ChatGPT (full history and memories), Spotify, Instagram, GitHub, YouTube, full Shopify shopping history, and more. Community-contributed connectors are landing regularly, and anyone can create one for any platform - it takes about 10 minutes with Claude Code.
Privacy by Design
Personal data is private for a reason. DataConnect is privacy-first by design: all scraping and processing happens client-side on the user's machine. Storage is user-controlled, with your local machine’s storage used by default. Data is permissioned by keys you control. It is bundled with a Vana personal server to make data portable.
Build a Data Connector
DataConnect is an open-source utility, and the community is building new connectors all the time. If you want to make your own data more portable, you can contribute a connector with a single AI agent prompt:
- Choose a platform where you have data you'd like to be able to port
- Use the connector framework and contribution guidelines in the repo to build a retrieval flow for your own account data
- Test the connector with your own account to verify that the data export works correctly
- Submit a PR for community review — if accepted, anyone can use it to exercise their own data portability rights
If you believe in portable, user-controlled data and want one tactical thing you can do this week to contribute: build a connector for a platform you personally use. Each new connector makes one more data source portable for everyone.
Build Apps with Portable Data
If you're building a consumer application or AI agent, you can start building with portable personal data today:
- Open the starter repo with your preferred vibe coding tool
- Tell the AI what you want: "Build me a Rick Rubin style application that surfaces poetic learnings from my ChatGPT conversations and Spotify listen history, and their relationship"
- Try it with your own data
- Ship to production, anyone who tries it will plug their own data in
There's a prompt in the repo that you can essentially one-shot an app with.
Imagine the design space when you have access to ten years of someone's cross-platform data, sleep history, conversations, messages, shopping patterns, and you can run privacy-preserving compute on top of it. That's what portable data unlocks.
Get Started
- Download DataConnect: account.vana.org/download-data-connect
- Build with portable data: github.com/vana-com/vana-connect-starter
- Create a data connector: github.com/vana-com/data-connectors
- Docs: docs.vana.org
In an AI-native world, portable data is what keeps the internet open, and what gives builders a competitive advantage in consumer AI products.
