# Verglas documentation

> Run cache-pathed lakehouse workloads with Verglas Cloud or a self-hosted Verglas server.

## Docs

- [Cloud quickstart](https://docs.verglas.dev/docs/get-started/cloud-quickstart.md): Log in to Verglas Cloud and deploy a scheduled worker from the CLI.
- [Self-host Verglas](https://docs.verglas.dev/docs/get-started/self-host.md): Run a single-node Verglas cache server with Docker Compose.
- [Verglas documentation](https://docs.verglas.dev/docs/index.md): Run Iceberg-aware data workloads in Verglas Cloud or on your own infrastructure.
- [Containers](https://docs.verglas.dev/docs/platform/containers.md): Deploy and operate long-lived tenant services in Verglas Cloud.
- [Follow local output](https://docs.verglas.dev/docs/platform/follow.md): Capture a local process or file as rows in an Iceberg table.
- [Workers](https://docs.verglas.dev/docs/platform/workers.md): Build bounded, event-driven workloads that run locally or in Verglas Cloud.
- [CLI reference](https://docs.verglas.dev/docs/reference/cli.md): Current Verglas command groups and common examples.
- [Worker specification](https://docs.verglas.dev/docs/reference/worker-spec.md): Reference for the portable JSON and TOML worker manifest.
- [SDK overview](https://docs.verglas.dev/docs/sdk/overview.md): Choose the TypeScript or Rust SDK and understand the shared data-plane contract.
- [Rust SDK](https://docs.verglas.dev/docs/sdk/rust.md): Use the native streaming client and worker contract from Rust.
- [TypeScript SDK](https://docs.verglas.dev/docs/sdk/typescript.md): Define workers and access tables, queues, graphs, feeds, and vector indexes from TypeScript.