Start with one interaction on a test website—for example, submitting a login form. By the end of this path, you should be able to find that interaction in Darwinium, check the data it contains and identify the deployed configuration that produced it.
Darwinium connects collection, journey configuration, risk assessment and an action in your application or at the edge. This first exercise establishes the collection and event path. You can then add and test the decision and action you need.
1. Choose the environment and interaction
Sign in to Darwinium Portal and select the node for your test environment using the selector at the top. A node determines the event data and configuration you are working with; see Nodes & environments.
Agree on one flow with your team. For this example, use an authentication journey with a step named login_submit. Identify the request's host, method and URL, and the account identifier you need to recognize returning users. Use test data.
You need access to the node, its workflow repository and the relevant deployment controls. Your administrator can assign these through Roles & permissions. Use the Portal tour to locate the controls.
2. Connect the application
For a website or API behind a supported CDN, start at the edge. This is our preferred integration route because you can extend protection across additional touch points through journey and route configuration.
Follow the setup guide for Cloudflare, AWS CloudFront or Akamai. Each guide specifies the provider permissions, credentials and target settings you need. Complete the setup for the test domain and environment you chose.
For other architectures, Install & collect explains JavaScript tags, mobile SDKs and direct API integration. Profiling collects device and behavior information; a configured CDN step or an Event API request supplies the interaction to assess. They are complementary parts of the integration.
3. Map the interaction to a journey step
Open Workflows. For a website, use Record your first journey: browse through the flow, review the discovered steps and update the journey configuration. You can also create a journey manually.
Open the resulting .journey.yaml file from Explorer to use the journey editor. Review the step that represents the login submission:
- Give it the name
login_submitand an appropriate event type, such asaccount_login. - Check that its host, method and URL match the request your website actually sends.
- Map the required account identifier from the request to the appropriate Darwinium attribute.
- Check which data is available from the request, response or earlier steps before using it in a rule.
See Create & configure steps and Dependencies & execution order. Recording discovers configuration for you to review; it does not deploy that configuration or decide your business policy.
4. Add profiling and any required integration
For browser device and behavioral data, follow Insert JavaScript profiling at the edge. Check both where the script is loaded and how its profiling data accompanies the protected interaction.
Use Darwinium Marketplace when the flow needs an additional service—for example, Turnstile verification or an identity-risk lookup. Marketplace packages supply reusable configuration; importing and configuring them is part of authoring the journey.
5. Validate, build and deploy
Follow Deployment confirmation. Confirm that the journey is included in journeys.yaml, the intended target and hosts are enabled, and the configuration builds successfully. Review and deploy the chosen revision to your test environment.
Saving a file changes the workspace. A successful build checks and prepares configuration. Deployment makes the selected revision active. Use Workflow editor & Git workflow for the full editing process.
6. Perform the flow and find the event
Visit the test website and perform the login submission. In the portal, open Investigations → Query, choose a time range that includes the test, and search for your configured names:
journey_name = "authentication" AND step_name = "login_submit"
Replace authentication with the journey's configured name. Select Search, then open the matching row. Confirm the step name, timestamp and mapped account identifier; check the profiling fields you enabled and any workflow outputs you configured. See Investigations overview.
If no event appears, check the selected node, time range, deployed revision and request trigger before changing rules. Troubleshoot installation covers the next checks. A step configured not to generate an event will not appear as an event row.
7. Connect assessment to action
Once the event path works, follow Build your first decision, then Apply decisions in your application. Decide whether the result will drive an edge action, an origin application response or an investigation workflow, and test that behavior explicitly.
Finally, record confirmed outcomes with Labels & feedback, and use Dashboards to monitor the traffic and decisions you have configured.