Question 1

What kind of work can I automate by connecting Dust to other tools?

Accepted Answer

You can automate workflows where Dust answers questions or generates responses based on your company data, and then route those outputs to other systems. Common patterns include having Dust process incoming requests and feed results into documentation tools, communication platforms, or databases, or triggering Dust to analyze data from connected sources on a schedule. The key is that Dust handles the intelligence layer, while connected tools handle the storage, notification, or follow-up actions.

Question 2

How does data move between Dust and other tools in an automation?

Accepted Answer

Data enters Dust through configured integrations with your company data sources (documents, databases, knowledge bases), and you can send Dust outputs to other tools via API or webhook connections. When automating, incoming triggers from connected tools pass information to Dust, which processes it against your data, and then you route the response back out to storage, communication, or action tools. You need to map which fields from incoming data feed into Dust's prompts and where the generated output should land in your downstream systems.

Question 3

What breaks or causes problems when automating with Dust?

Accepted Answer

Dust relies on the quality and freshness of your connected data sources, so outdated or incomplete information in those sources will produce unreliable outputs. If your automation doesn't clearly specify which data context Dust should reference, it may generate responses based on incomplete information. Also, Dust's output format and length can vary, which may require parsing or transformation before sending to downstream tools. Finally, if you don't properly handle cases where Dust cannot find relevant information, automations may fail or produce generic responses instead of triggering alternative workflows.

Question 4

How should I set up Dust so automations run reliably?

Accepted Answer

Start by clearly defining which of your company data sources Dust should access, and verify those sources are kept current. Structure your Dust assistant with explicit instructions about what data to search and how to format responses for the downstream tools you're connecting. Test the end-to-end automation with real data before going live, and set up monitoring to catch cases where Dust cannot answer or where outputs are malformed. Document the exact data fields and contexts each automation expects so updates to your data sources don't break the flow.

Automate the manual work in Dust

Dust by the numbers

Top 5 Dust friction points when done manually

Dust conversations scattered across channels

Manual context switching between Dust and work

Dust responses require manual verification

Onboarding new team members into Dust

Dust insights require manual follow-up actions

What you can automate in Dust

Manage new or updated Google Sheets rows by talking to Dust assistants

Upload documents in Dust whenever new events are added in Google Calendar

Upload documents to Dust from new or updated rows in Google Sheets

Upload documents to Dust when new contacts are recently created or updated in HubSpot

Upload new Fathom transcripts as documents in Dust

Upload new Gmail emails to a Dust data source folder

Tools commonly connected to Dust

Dust Automation questions

Ready to automate your work in Dust?