From Database Changes to Useful Business Updates

A colleague updates an expense record. The person preparing the project budget does not notice until the next meeting. Nobody made a mistake; the information changed, but the process around it never moved.

That gap is a useful place to start with automation. In Chatigram, a database operation can trigger a workflow when a record is created, updated or deleted, according to the operations you select. The workflow can then connect that event to a message, another application or further processing.

The aim is a useful next step. Sending an email for every edit usually creates more noise than clarity.

Choose one change that matters

Start with a process you can explain in a sentence: “When a new expense is recorded, let the project coordinator know which project it belongs to.”

Decide what the recipient needs to do after reading the update. If there is no action, a scheduled summary may be more appropriate. If timing matters, an event-triggered notice may be worth building.

For this example, assume a small team maintains a Projects table and an Expenses table. An expense includes a name, an amount and a project reference. This is an illustrative setup you can adapt, not a required Chatigram template.

Make the records understandable first

The database editor gives you a place to work with records, columns and filters. Use suitable field types and consistent names. An amount should be usable as a number; a project reference should identify the intended project reliably.

Before connecting automation, enter a few realistic sample records. Include one incomplete record and one corrected entry. Decide whether a missing project reference should prevent a notification or produce a request for more information.

Chatigram databases also provide API access for connected applications. Review the generated REST API and OpenAPI specification when another system needs to read or write your data. Keep credentials in the appropriate configuration rather than putting them into message text.

See Databases on the platform page for the broader relationship between records, workflows and applications.

Build the smallest complete workflow

Create a workflow using the Database Operation trigger. Select the database, the Expenses table and the operations relevant to your process. For a first version, a created-record notice is easier to reason about than a notification for every possible change.

A simple flow could connect the trigger to a Gmail message action. Select the intended account connection and recipient. Include the expense name, amount and project reference from the available event or record data, checking the actual output in your workspace.

Chatigram supports multiple connections for supported third-party services. If you have a personal mailbox and a business mailbox connected, choose deliberately which one sends this update.

Test with your own inbox first. Check the values and the meaning of the message before including other people.

Add business rules before adding AI

An expense amount does not require AI to decide whether it exceeds a threshold. A workflow condition can represent the rule you choose. AI may be helpful later for summarizing descriptions, but the underlying amounts and project references should remain grounded in the stored records.

For example, you might want an immediate notice only for a new expense above your chosen amount. Smaller expenses could be included in a separate weekly report started by Cron. These are two different workflows with different jobs.

For a summary, calculate or retrieve the totals from the records and have AI explain the results. Check that the reporting period, currency and included records match the intended calculation. A polished paragraph is not evidence that the numbers are correct.

Plan for repeated events and failed actions

A record may be edited several times. Decide whether each update deserves another message. Where your workflow design requires duplicate protection, use a stable record identifier and track whether the relevant action has already been completed.

Be careful when the workflow writes back to the same table that triggers it. That write may create another qualifying event. Review the selected operations and add conditions that distinguish the original change from your workflow’s own updates.

Test a disconnected email account or a failed API action. Decide how you will recognize a failure and recover it without sending the same notice repeatedly. The appropriate handling depends on the actions and logic you configure.

Extend the pattern when the first version is useful

The same approach can support project changes or an application requesting fresh data through an HTTP endpoint. You can also connect external software through webhooks and API nodes, depending on that software’s available API.

Keep the purpose visible: what changed, who needs to know, and what should happen next. Review a week’s worth of notices with the recipient before adding more triggers.

Try the interactive examples above, then explore the workflow cookbook or business solutions. The pricing page shows which plans include databases, so you can choose capacity around the process you intend to build.

• automation
• workflows
• business productivity