Applying Agile Principles in Azure DevOps: Team Management, Planning, and Process Improvements

This article, presented by Aaron Bjork, describes how the Azure DevOps and Azure DevOps Server teams apply agile principles to manage teams, roles, planning, iterations, and process improvements.

Azure DevOps is a subscription‑based cloud service with about 800 employees that runs three‑week iterations and releases at the end of each iteration. Azure DevOps Server is the on‑premises counterpart, packaged three to four times per year and deployed via update bundles.

The introduction of agile is illustrated with a before‑and‑after comparison (see image).

Four key changes were made:

Culture change – focusing on autonomy, mastery, and purpose, and balancing alignment with autonomy.

Team attitude shift – moving from individual focus to cross‑functional, self‑managed teams of 10‑12 members with clear charters, independent workspaces, and ownership of features and deployments.

Change in planning and learning – adopting a four‑level planning hierarchy (iteration, quarterly, semi‑annual, strategic) and encouraging rapid feedback and deployment.

New health‑preserving methods – implementing a defect‑limit rule (engineer count × 5) and shielding interference by splitting each sprint into a feature team (F) and a barrier team (L) that handles production incidents.

Team autonomy is reinforced through a “yellow sticky” movement every ~18 months, allowing developers to choose the product area they wish to work on for the next two planning cycles; about 80 % stay with their current team.

Transparency and accountability are ensured by having each team send an email at the end of every sprint summarizing completed work and upcoming plans.

Key takeaways include treating agile seriously but not dogmatically, celebrating results rather than activities, adopting a mindset that every iteration must release a new service or update, and recognizing that the desired culture drives the desired behavior.