Workflow Concepts
Workflows are how XY turns repeatable operational work into automation. A workflow can begin with a browser recording, a file, a connected app, or a business event, then move that work through the right steps until the outcome is complete.
What a workflow can include
Most XY workflows combine some or all of these ingredients:
Agents
Agents help with specialized work such as extracting information, checking completeness, or answering questions from trusted files.
Integrations
Integrations connect XY to other systems so a workflow can read data, send updates, or pull in documents.
Files
Files provide both input and context. A workflow might process an uploaded document, or it might use policies and SOPs as reference material.
Browser automation
When work happens inside a portal or website, a workflow can use the XY Browser Agent and recorded browser steps to complete that part of the process.
Human review
Not every case should run straight through. Workflows can send exceptions to a person when something needs review or confirmation.
Two workflow starting points
Configured workflows
These are workflows your team builds or manages directly in the web app.
Recorded workflows
These begin in the Browser Agent, where a user records a real task and saves it for reuse.
A typical workflow lifecycle
- A team creates or records the workflow.
- The workflow is tested on a safe example.
- The team refines the steps, inputs, and connected systems.
- The workflow is reused for repeated work.
- Exceptions are reviewed by a person when needed.
Common workflow patterns
File-driven
A file is uploaded or synced into XY, then a workflow reads it, extracts the key information, and routes the result to the right next step.
Integration-driven
A workflow starts from a connected system and uses that connection to move information across tools.
Browser-driven
A workflow uses recorded browser steps when the work has to happen inside a website or portal.
Hybrid
Many of the strongest workflows combine all three. For example, a workflow might read a file, look up supporting data through an integration, and then finish the last steps in a browser portal.
How teams usually roll workflows out
Start small
Pick one narrow process with a clear outcome.
Keep a person involved early
Use a co-pilot or supervised approach while your team learns what the workflow does well and where it still needs review.
Expand after confidence grows
Once the workflow is reliable, connect more inputs, add more users, and apply the same pattern to adjacent processes.
Example
Here is a simple XY workflow:
- A referral document is uploaded into
Files. - An agent extracts the key details.
- The workflow checks the right internal policy or SOP.
- A browser workflow completes the next steps in a portal.
- A staff member reviews any exception before final submission.
Why workflows matter
Workflows help teams:
- reduce repetitive manual effort
- make processes more consistent
- reuse the same automation across many items
- combine files, apps, and browser work in one place
- keep people involved where judgment matters