Skip to main content
Step 1 of 4. Connect a source or upload a file so Addison has data to work with. If your project already has the files or tables you need, skip ahead to Create your first analysis.

Bring data in

Click Connectors in the left sidebar, then New connection at the top right. The wizard has three steps:
The Connectors page titled 'Connectors — Create and manage data connections and app integrations', with a New connection button in the top-right and Data / Apps tabs below. A table lists connections by Name and Status, each with its source type beneath the name: bq_test (BigQuery), DEMO_WH (Snowflake), shortify-url (GitHub), Salesforce connection fresh (Salesforce), Shopify connection (Shopify), and NetSuite connection (NetSuite), each with a 'Last updated' date and a status dot.

The Connectors page: your data connections and their status

1

Choose a data source

Snowflake, Postgres, and BigQuery sit under Popular; the rest are under Other.
Step 1 of 3 of the New connection wizard, titled 'Choose a data source' with a progress indicator reading 1/3. A Popular row offers Snowflake (Data warehouse), Postgres (Database), and BigQuery (Data warehouse). An Other section lists S3 (File storage), GitHub (Repository), Databricks (Lakehouse), ClickHouse, MySQL, Oracle, MotherDuck, Redshift, GCS, SQL Server, REST API (HTTP endpoint), MongoDB (Document database), and Apache Iceberg (Lakehouse catalog). An Apps section below starts with Shopify, NetSuite, and Salesforce.

Step 1: choose a data source

2

Enter the details

Give the connection a Name. It’s the identifier used in SQL later, so use letters, numbers, hyphens, and underscores only. Then fill in the credentials and click Test connection before continuing. A pass means Summation could authenticate and reach your data.
Step 2 of 3 of the wizard, titled 'Connection details' (2/3). It has Name and optional Description fields, a 'Need help setting this up? View the setup guide' link, and a Snowflake section with an Authentication toggle between Password and Key pair, plus Account Identifier (myorg-myaccount), Warehouse (COMPUTE_WH), optional Role (ACCOUNTADMIN), Username, and Password fields. Below is a 'Scheduled snapshots' row with an Enabled checkbox and the note 'Datasets inherit this schedule unless they override or disable it.', and Test connection, Back, and Continue buttons with the hint 'Test your connection before continuing.'

Step 2: connection details and credentials

3

Add datasets

Pick the tables to bring in. Each one becomes a dataset available to Addison, to artifacts, and to your tables and data catalog.
An 'Add datasets to this connector' dialog with a checkbox tree on the left rooted at '/' showing DEMO_DB, SNOWFLAKE, SNOWFLAKE_LEARNING_DB, and SNOWFLAKE_SAMPLE_DATA, and a searchable list on the right showing the same four entries each tagged with a yellow DATABASE badge and a drill-in chevron, above a footer reading '0 datasets selected' with Cancel and Add dataset buttons.

Step 3: pick the datasets to bring in

How Summation reads and refreshes data depends on the connector and its settings. See Connectors for details.

Check that it landed

A connection puts its datasets under Tables in the left sidebar; click one to see the rows. An uploaded file shows up in its project’s file list instead, and appears under Tables only once you have asked Addison to convert it.
The Tables page titled 'Create and manage your tables', with a Create button top-right. Type filter chips read All, Summation (21), and Snowflake (8). A table lists folders and tables with Name, Type, Column, Source, and Last refreshed columns: a Sales_Data folder, then tables allocate_plan, budget_plan, buy_plan, campaign_performance, and commissions_actuals, each sourced from Summation and last refreshed Aug 31.

The Tables page

Projects keep the files and context for a task together. Attach the tables and views you need as the project’s key tables so Addison reaches for them first here. Addison can also work with uploaded project files without converting them to tables first.
Start narrow. Choose the data needed for one question. You can add business context as you review the answer.

Next

Create your first analysis

Step 2: ask a question, create a report, and check its evidence.