Data Tables
Data Tables hold structured, row-based data that your agents can query and display. They generalize beyond product catalogs to support any type of structured data — ad campaign metrics, rate cards, audience segments, or custom datasets.
Table types
| Type | Use case |
|---|---|
| Product Catalog | Products, plans, rooms, menu items — anything with a title, price, image, and CTA |
| Rate Card | Pricing tiers, service rates, subscription plans |
| Audience Segments | Customer segments, cohorts, targeting groups |
| Campaign Data | Ad campaign structure and performance data (synced from Meta, Google, TikTok) |
| Custom | Any structured data that doesn't fit the above categories |
Creating a Data Table
Go to Knowledge & Data in the left nav, then switch to the Data Tables section. Click New Table.
- Choose a table type from the list above
- Give it a name and optional description
- Choose a category (ecommerce, finance, travel, SaaS, food, real estate, education, or generic)
- Define your column schema — each column has a key, label, and type
Adding data
CSV / Excel upload
Upload a CSV or Excel file and map its columns to your table schema. For large files, Zimmer uses presigned S3 URLs so uploads go directly to cloud storage without hitting server size limits.
- Click Upload on your Data Table
- Select your file
- Map source columns to table columns
- Review the preview and confirm
API feed
Connect an external REST API endpoint that Zimmer polls on a schedule:
- Click API Feed on your Data Table
- Enter the endpoint URL, HTTP method, headers, and the JSON response path where records live
- Set a sync schedule: manual, hourly, daily, or weekly
- Zimmer fetches and upserts records automatically on each sync
API feeds are useful for keeping data in sync with Shopify product APIs, inventory systems, pricing engines, or any service with a REST endpoint.
Manual entry
Add records one at a time through the UI. Best for small, rarely-changing datasets.
Integration sync
For ad platforms (Meta Ads, Google Ads, TikTok Ads, DV360), campaign data syncs into Data Tables automatically when you connect the integration. You do not need to create these tables manually — they are created and updated by the sync process.
Linking to agents
Go to your agent's Knowledge tab and select the Data Tables this agent can access. For Knowledge agents, the AI uses table data to answer questions and display product cards. For Action agents, table data provides context for optimization decisions.
Using in journeys
Add a Product Card node to your journey flow. Configure it to pull from a linked Data Table:
- Fixed item — Always show a specific record
- Category filter — Show items matching a category
- AI recommendation — Let the AI pick the most relevant item based on conversation context
Schema and columns
Each Data Table has a column schema that defines its structure. Columns have:
- Key — Internal identifier (used in API feeds and CSV mapping)
- Label — Display name shown in the UI
- Type — Text, number, URL, image URL, boolean, date
- Required — Whether the column must have a value
You can update the schema after creation. Existing records retain their data; new columns start empty for existing records.
Sync status
Each Data Table tracks:
- Record count — Total active records
- Source type — CSV Upload, Manual, or API Feed
- Last synced — When the data was last updated
- Sync schedule — How often API feeds refresh