Project Deep Dive

Insights from the PM

Service Groups and Relationships

As the Lead Product Manager for Azure Resource Manager, I led the strategy, design, and delivery of Azure Service Groups—a major evolution in how Azure organizes and governs cloud resources. This initiative introduced a graph-based relationship model to overcome the limitations of Azure’s hierarchical Management Groups, enabling many-to-many resource connections, dynamic group memberships, and cross-boundary flexibility.

The Challenge

Azure’s native Management Group hierarchy enforced a one-parent-per-subscription tree model. While scalable and simple, this structure proved too rigid for customers with complex organizational, regulatory, or deployment needs. Multiple enterprise customers—including top financial and government clients—requested more expressive ways to organize their cloud assets across teams, services, and environments.

At the same time, internal Azure engineering teams were independently building custom grouping solutions, creating fragmentation across the platform. We needed a unified, extensible solution that could serve both customer demands and internal partner needs.

My Role

As the lead PM, I was responsible for:

  • Product Strategy: Defining the next-generation grouping model for Azure

  • Cross-Team Alignment: Orchestrating a roadmap across 12+ partner engineering teams

  • Architecture Decisions: Shaping how links, membership, and relationships would work

  • Delivery: Driving the feature from concept through public release under tight deadlines

The Solution

We introduced Azure Service Groups, a new platform primitive that supports:

  • Graph-Based Relationships: Many-to-many links between resources, groups, and services

  • Dynamic Memberships: Rule-based logic for automatically updating group contents

  • Extensible Design: A shared, persisted relationship model usable by all Azure services

To support this, I worked with engineering to build a robust linking mechanism, using persisted metadata rather than ephemeral queries. We scoped dynamic rules for future iterations while focusing the MVP on persisted links to accelerate adoption.

I also defined six guiding product principles—covering scale, performance, governance, and extensibility—which served as alignment tools for all partner teams.

Execution Highlights

  • Partner Collaboration: Engaged 12+ Azure teams including Identity, Compute, Networking, and Security to adopt the Service Group platform

  • Engineering Negotiations: When a key partner (Identity) lacked capacity, I brokered a deal to embed one of our principal engineers into their codebase, saving six months of delay

  • Deadline Focus: Established daily syncs, backlog forecasting, and stakeholder dashboards to meet an aggressive Q4 production readiness deadline, ahead of Azure’s holiday deployment freeze

Results

  • Launched in early 2025 with documentation, partner onboarding, and Microsoft Build coverage

  • Adoption: Partner services began actively integrating SGs into their customer experiences

  • Impact: Internal analysis surfaced over 2 million relationships across Azure services in just the initial inventory phase—highlighting the value and need for graph-based governance

  • Strategic Foundation: Positioned Service Groups as the future of Azure resource modeling, policy evaluation, and compliance mapping

Reflection

Service Groups was one of the most ambitious and rewarding initiatives of my time at Microsoft. It combined deep platform strategy, technical design, and cross-organizational leadership—pushing Azure toward a more flexible and intelligent governance model. Although my tenure was cut short by broader organizational restructuring, the project’s success and early traction underscored the long-term value of the investment.

Additional images and concepts

Previous
Previous

Management Groups

Next
Next

Tag Modernization