Skip to main content

Integrations

Q247 does not generate data on its own. Everything visible on the dashboards comes from systems that already run at the Client: code repositories, Jira, Confluence, the identity provider. This part of Configuration handles the connections to them.

Integrations fall into three groups of a different character. The first delivers work data, and without it the platform is empty. The second concerns accounts and sign-in and is optional, because the organization can manage accounts manually. The third supports AI functions: it accepts data about the use of AI tools and provides the language model to the AI Assistant.

Integrations that deliver data​

IntegrationWhat it bringsWhen it is needed
Connectorscommits and repository history, from which Calories, Increments and Lines are calculatedalways, this is the foundation of how Q247 works
Jirathe history of ticket status changeswhen the organization wants to measure Lead Time and Cycle Time
Confluencethe effort put into documentationwhen documentation is a significant part of the team's work

Connectors are a subcategory: one connector corresponds to one version control system, and Q247 supports seven types, from GitLab and GitHub to repositories added manually.

Running code analysis is a guide to the whole process, from creating the connector to data appearing on the dashboard.

The starting point for a deployment from scratch

The order of steps matters more than the details of individual fields. The symptom of skipping one of the steps looks identical to an integration failure: the connector shows an Operational status, and the dashboards are empty.

Identity integrations​

These two handle different things and are deployed independently.

IntegrationWhat it handlesWhat it does not do
SSO (OIDC)sign-in through an external identity providerdoes not create accounts
SCIM Provisioningcreating, updating and deactivating accounts and teamshas no effect on the sign-in method

You can deploy one of them, both, or neither. With both at once, the identity provider creates the accounts, and users sign in without a password on the Q247 side. With SSO alone, accounts have to be created manually, otherwise sign-in will not find their counterpart in Q247.

AI feature integrations​

Both are optional and independent of each other, and each supports a different AI function.

IntegrationWhat it handlesData direction
AI tool logsreceiving telemetry from AI tools, including Claude Codeto Q247
Language Modelthe model powering the AI Assistantfrom Q247 to the model

The first feeds the numbers in AI adoption, the second is responsible for the responses of the AI Assistant. AI adoption also works without telemetry, relying then on commit message trailers.

Actions outside Q247​

Integrations are the only part of Configuration that requires actions in other systems. Depending on the scope of the deployment, you will need help from people with the right permissions:

  • the version control system administrator: a service account and a token with read access to the repositories covered by the analysis,
  • the Jira or Confluence administrator: once, to install the Q247 app (cloud versions) or register the webhook (Data Center),
  • the identity provider administrator: registering the app for SSO or configuring the SCIM connector,
  • the network team: for deployments in which the plugin runs in the Client's infrastructure.

Every integration page has its own prerequisites section with a specific list.

Subpages​

See also​