> ## Documentation Index
> Fetch the complete documentation index at: https://docs.arcyria.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Getting started

> Connect a source and destination, configure the tabbed pipeline workspace, publish SQL, and run a supported pipeline.

The current app flow is connection-first and workspace-driven. You create
reusable connections and a lightweight pipeline, then complete setup on the
pipeline detail page. There is no canvas editor.

## Current pipeline workspace

Every pipeline opens at `/workspace/data-pipelines/[id]` with these tabs:

| Tab                 | Purpose                                                                                                            |
| ------------------- | ------------------------------------------------------------------------------------------------------------------ |
| **Overview**        | Review the source → pipeline → transformations → destinations relationship and readiness checks                    |
| **Source**          | Test the source, discover the catalog, select streams, choose sync modes/cursors, inspect fields, and preview rows |
| **Transformations** | Manage destination-owned SQL models, drafts, published revisions, and revision history                             |
| **Destinations**    | Add and edit independent delivery targets and their output assignments                                             |
| **Runs**            | Inspect executions, rows, revisions, destination targets, warnings, and failures                                   |
| **GitHub**          | Manage the pipeline's repository integration when enabled                                                          |
| **Settings**        | Edit metadata, activate/pause, validate, schedule, or delete the pipeline                                          |

## Current app flow

1. Create a PostgreSQL, MySQL, MongoDB, Airtable, Asana, Notion, HubSpot, or Stripe source connection.
2. Create a PostgreSQL, MySQL, MongoDB, or Airtable destination connection.
3. Create a pipeline from **Data Pipelines** with a name and source connection.
4. Open the pipeline and use **Source** to discover, preview, select, and save streams.
5. Use **Destinations** → **Add destination** to choose a connection and save a delivery target.
6. Use **Transformations**, select that destination, and create a SQL transformation.
7. Save the draft, then **Validate draft**, **Preview output**, and **Publish revision**.
8. Return to the destination editor, configure the published output table and Upsert keys, then save.
9. Use **Overview** → **Validate pipeline** until all readiness checks pass.
10. Click **Run all** in the pipeline header.
11. Use **Runs** to inspect the execution.
12. Add a fixed interval or UTC cron schedule from **Settings** after a manual run succeeds.

## Before you configure delivery

The SQL delivery runner does not create final destination tables. For database
destinations, use **Create or verify destination table** when that explicit
setup action is available, or create the reviewed DDL yourself. For Airtable,
create the table and writable fields in Airtable first, then refresh them in the
destination editor. MongoDB can create a collection on first write, but
production collections should have reviewed validators and Upsert-key indexes.

## Next

Follow the [Quick Start](/getting-started/quick-start) for an end-to-end PostgreSQL example.
For CRM data, follow the [HubSpot source guide](/connections/sources/crm/hubspot).
For bidirectional SaaS/database delivery, follow the
[Airtable and MySQL guide](/example/pipelines/airtable-and-mysql).
For document workflows, follow the
[MongoDB and PostgreSQL guide](/example/pipelines/mongodb-and-postgresql).