You probably have notes in Apple Notes, tasks in Todoist, contacts in Google Contacts, bookmarks in Raindrop, and files in Google Drive. Each of these tools stores your data on their servers, in their format, under their terms of service.

A data pod is the alternative: a dedicated database — specifically, a PostgreSQL server — that holds all your personal data in one place. You own it. You control who accesses it. You can export it, query it, or move it at any time.

Technically, a data pod is a provisioned PostgreSQL instance with a defined schema for personal data: entities (notes, tasks, contacts, bookmarks, files), relationships between them, views, channels, and AI interaction history. It runs on a server — either one we provision for you in the EU or US, or one you self-host on your own infrastructure.

It is not a cloud storage bucket. It is not a folder of markdown files. It is a real relational database with typed tables, foreign key constraints, full-text search indexes, and vector embeddings for semantic search. Standard SQL. Standard tooling. No proprietary formats.

Three reasons: portability, AI access, and resilience.

Portability. When your data lives in five SaaS products, each one holds your data hostage. The "export" from most tools is a ZIP of markdown files that loses 80% of the structure — no relationships, no properties, no metadata. With a data pod, you run pg_dump and get everything: the complete schema, all data, all relationships. Import it into any PostgreSQL-compatible system anywhere.

AI access. AI can only help you with data it can see. When your information is scattered across five services, no AI can reason about all of it unless you manually paste context into a chat window. A data pod gives AI structured access to your entire knowledge base through a typed API. "Find all contacts related to this project" becomes a real query, not a guessing game.

Resilience. SaaS companies change pricing, get acquired, pivot, or shut down. Every year, some tool you rely on makes a breaking change. When your data lives on a pod you own, these events are inconveniences, not catastrophes. The data is already on your server. Worst case, you switch clients.

When you sign up for Synap, we provision a dedicated PostgreSQL instance for you. Not a shared database with row-level security — a separate server that only your account can access. Your entities, relationships, AI history, and files live there.

The Synap desktop app connects to your pod. AI agents connect to your pod through the Hub Protocol. Connectors pull data from external services (Google Calendar, GitHub) into your pod. Everything converges on one database you own.

If you prefer to self-host, you can run the pod on your own infrastructure. Docker image, standard PostgreSQL, connect the same desktop app. The architecture is the same either way — the only difference is who manages the server.

We are entering an era where personal infrastructure makes sense again. Cloud storage commoditized. Compute is cheap. AI needs structured data access to be useful. The missing piece was a data layer designed for personal use — not enterprise SaaS scaled down, but personal-first infrastructure that can scale up.

That is what a data pod is. Your personal server. A place where your digital life lives in a format you control, accessible to any AI you choose, portable to any platform, and yours permanently.