Connecting Your Data

Overview

Stackless connects business systems to a customer-specific data warehouse, then makes those source tables available for transformations, semantic models, dashboards, and Agent workflows.

The exact setup flow depends on the source. Some sources use an authorization card in Stackless. Some require an OAuth handoff. Some require the Stackless team to configure a database, API, file drop, or private credential path.

[picture 1. Sources page with connector status, setup state, and sync freshness]

Common source categories

Stackless supports source categories such as:

• Point-of-sale and commerce systems.
• Payment, billing, and accounting systems.
• CRM, support, and operations systems.
• Marketing and advertising platforms.
• Databases, spreadsheets, file exports, and custom APIs.

This list is not exhaustive. The Sources page in your Stackless workspace shows the source types and connectors available for your environment.

Connection lifecycle

1. Request or start the source

Depending on your permissions, you may be able to start source setup from Data > Sources or through the Agent. If setup requires credentials or an external approval, Stackless will guide the authorized user through that step.

2. Authorize read access

Most analytical sources only require read access. For OAuth-based tools, the owner signs into the source system and grants Stackless access. For database or API sources, Stackless will provide the secure credential requirements.

3. Initial sync

After authorization, Stackless runs an initial sync into the warehouse. Historical sync time depends on the source API, data volume, and source rate limits.

4. Catalog refresh

After source tables are available, the Catalog refreshes so the Agent and data pages can discover the new assets. If a source is newly connected and the Agent cannot see it yet, ask an admin or Stackless contact whether the Catalog refresh has completed.

5. Transformation and modeling

Raw source tables are usually not the final reporting layer. Stackless builds Transformation Models to clean and combine the data, then Semantic Models to define measures and dimensions for dashboards and Agent charting.

Asking the Agent about sources

Useful prompts:

• "What sources are connected in this workspace?"
• "Which source tables were refreshed recently?"
• "Do we have order, customer, and ad spend data available?"
• "Find the table that contains campaign spend."
• "Is this connector healthy?"

If the Agent says a source is unavailable, distinguish three cases:

• The source is not connected.
• The source exists, but your role does not have access to its schema.
• The source was connected recently, but Catalog metadata has not refreshed yet.

Data freshness

Freshness varies by source and configuration. Some sources sync frequently throughout the day; others sync daily or on a schedule that matches the upstream system. Use the Sources page, Catalog freshness, dashboards, or Monitor to understand current freshness.

Uploading files

For ad hoc CSV or XLSX files, use the upload button in the Agent. Tell the Agent whether the upload is:

• Seed data for a new analysis.
• A one-time file to compare against warehouse data.
• A file that should be matched to an existing source.

See Uploads for the full workflow.

What happens next

Once data is connected, the normal Stackless path is:

1. Source data lands in the warehouse.
2. Transformation Models clean and reshape it.
3. Semantic Models define measures, dimensions, and joins.
4. Dashboards and Agent artifacts use those governed models.
5. Monitor tracks recurring runs and failures.