Skip to main content

Key concepts in SEI 2.0

Last updated on

Overview​

Harness Software Engineering Insights (SEI) 2.0 is a complete re-imagination of how engineering organizations can understand, measure, and improve software delivery. It brings together signals across your software delivery lifecycle (SDLC), connects them with organizational context, and transforms them into actionable insights for teams, managers, and executives.

SEI 2.0 is built around a set of flexible building blocks that help you answer critical questions like:

  • Are our engineering teams operating efficiently?
  • How is work flowing from idea to production?
  • Where are the bottlenecks in our delivery pipeline?
  • Are we meeting our reliability and velocity goals?

Core Concepts​

To answer these questions, SEI 2.0 introduces a simplified and extensible data model.


This data model diagram walks through how SEI handles organizational data and insights, including:

  • Developers: Imported from HR systems via CSV and managed at the account level.
  • Org Tree(s): Developers are grouped based on attributes like Manager, Location, Role, Employment Type, or Project. Leaf nods of the Org Tree becomes Teams.
  • Team: Team Managers are assigned by the SEI Admin at the project level and configure team-specific settings (e.g. Jira projects, Git repos, production services, and pipelines) influenced by associated Profiles.
  • Profiles: Define which insights to enable and how they are computed, and are associated with one or more Org Trees to guide all insights calculations.
  • Dashboards in Insights: The primary way insights are consumed in SEI 2.0. Dashboards are generated using data filtered by Profiles, Teams, and Org Tree context, and are viewed on the Insights page.
tip

The foundation of Harness SEI is the account-level Integrations setup by the SEI Admin. These integrations must be configured before importing developers or creating Org Trees.

Developers​

Developers are the foundational unit in SEI 2.0. Every insight, metric, and trend in the platform starts with accurate identification and attribution of developer activity across your toolchain. SEI automatically handles the complexity of identity mapping across systems like GitHub, GitLab, Jira, Bitbucket, PagerDuty, and more so you don’t have to based on developer email. Developers can belong to multiple teams across different Org Trees and Projects, based on the grouping logic defined in your Org Tree.

Org Tree​

The Org Tree represents your organizational hierarchy and structure in SEI 2.0. It provides the foundation for grouping developers and replicating the org hierarchy into SEI 2.0. At present, the Org Tree is powered by CSV-based imports from your HRIS system (e.g., Workday, BambooHR). The CSV must include a contributor attribute (such as ManagerEmail or ManagerID) that defines the reporting relationship between individuals.

Using this structure, SEI 2.0 automatically builds a manager and reportee tree, allowing you to view and analyze metrics in the context of real-world teams and reporting lines.

Multi-level groupings based on custom attributes (e.g., Department → Function → Team) are supported in SEI 2.0. This allows organizations to model more complex hierarchies and operational structures, making it easier to analyze engineering performance across departments, locations, and lines of business.

Teams​

Teams are the core unit of measurement in SEI 2.0. Every leaf node in the Org Tree is treated as a Team, making it the fundamental grouping for surfacing insights, applying goals, and driving accountability.

Each Team represents a group of developers working together within the organizational hierarchy and is automatically derived from the Org Tree structure.

Key Characteristics​

  • Auto-derived: Every leaf node in the Org Tree is automatically considered a Team.
  • Configurable: Each Team has its own configuration to define how metrics are calculated and displayed.
  • Contextualized and metric specific settings: Teams include metadata such as associated integrations, relevant services, repositories, destination branches, pipelines, environments, etc.

This configuration layer adds precision to how metrics like Deployment Frequency, Lead Time, MTTR, and others are calculated, which ensures each metric reflects the real scope, velocity, and complexity of the team's delivery lifecycle.

Profiles​

Profiles control what SEI 2.0 measures and how it measures it by enabling or disabling specific metrics and defining which event types power those metrics. They provide a flexible way to customize SEI 2.0 insights to your organization’s goals and workflows.

Each Profile includes which metrics are enabled or disabled (e.g., Deployment Frequency, Lead Time, Change Failure Rate) and the events and signals that power those metrics.

Types of Profiles​

SEI 2.0 supports three main types of Profiles, each optimized for different measurement objectives:

Type of profilePurposeMetrics
EfficiencyDelivery performance (DORA metrics) & sprint insightsDeployment Frequency, Lead Time for Changes, Change Failure Rate, MTTR
ProductivityDeveloper productivity and collaborationPR Velocity per Dev, PR Cycle Time, Work Completed Per Dev, Coding Days Per Dev, Number of Comments Per PR, Average Time to First Comment
Business AlignmentAlignment of engineering output to business goalsBusiness Alignment

Relationship to Org Tree & Teams​

  • Teams define where the data comes from (repositories, pipelines, services, etc).
  • Profiles define what data is used and how to calculate each metric.
  • Each Org Tree can have only one Profile of each type (Efficiency, Productivity, Business Alignment).
  • The assigned Profile applies to the entire Org Tree and all Teams within it.
  • This ensures consistent measurement and reporting across all organizational units under that tree.

Together, Developers, Org Trees, Teams, and Profiles define how data is collected, grouped, and calculated in SEI 2.0. This data is surfaced in dashboards on the Insights page.

Dashboards in Insights​

SEI 2.0 surfaces all metrics and insights through dashboards on the Insights page, where dashboards are viewed and shared. These include both out-of-the-box dashboards and custom dashboards created by your team.

Whether you're focused on shipping faster, improving collaboration, or aligning engineering with business goals, the Insights page is your central place to track performance.

Out-of-the-box dashboards​

SEI 2.0 includes the following pre-built dashboards designed to deliver value from day one:

  • Efficiency: Get a complete picture of your engineering throughput and stability. This dashboard combines DORA Metrics and Sprint Insights to highlight how reliably and predictably your teams are delivering

    MetricDescription
    Deployment FrequencyHow often code is successfully deployed to production.
    Lead Time for ChangesTime taken for code to go from commit to production.
    MTTRMean Time to Recovery after a failure in production.
    Change Failure RatePercentage of deployments that result in a failure in production.

  • Productivity: Understand how developers are working across pull requests, reviews, commits, and collaboration. Built for visibility into the day-to-day collaboration and flow of your teams.

    MetricDescription
    PR Velocity per DevNumber of pull requests completed per developer.
    PR Cycle TimeTime taken for a pull request to go from open to merged.
    Work Completed Per DevAmount of work completed per developer over time.
    Coding Days Per DevNumber of days a developer actively contributes code.
    Number of Comments Per PRAverage number of comments on pull requests.
    Average Time to First CommentTime taken to receive the first comment on a pull request.

  • Business Alignment: The one dashboard that you need to track how engineering output maps to product and business goals.

    Built for Every Level of the Org

    All dashboards are automatically scoped using your Org Tree, Teams, and Profiles so each viewer (from developer, team lead, to executives) sees the insights most relevant to them.

Canvas dashboards​

tip

Canvas dashboards are in beta. To request access, contact Harness Support.

If the out-of-the-box dashboards on the Insights page don’t meet your needs, you can use Canvas to create custom dashboards in SEI 2.0.

Canvas is where dashboards are created and managed. Once created, they are published to the Insights page, where they can be viewed and shared across your team.