Supergraph Architecture Framework

The Apollo Supergraph Architecture Framework (SAF) provides a collection of best practices in supergraph development. These practices stem from Apollo's extensive experience collaborating with numerous engineering teams in building their supergraph implementations.

The SAF includes an assessment to quantify your supergraph's current state and identify areas for improvement. Whether you're just starting with a supergraph or have been in production for years, you can use the SAF assessment to evaluate how well your API strategy aligns with Apollo's proven best practices.

The results of the SAF assessment serve as a roadmap for architecture and platform teams, guiding your organization toward an optimal supergraph implementation.

SAF pillars

SAF centers its principles and best practices around five pillars:

• Operational excellence: Efficiently operate, monitor, and manage your supergraph.
• Security: Protect your supergraph from intentional and unintentional misuse and breaches.
• Reliability: Design your supergraph for scale and high availability.
• Performance: Improve your supergraph's request and response performance.
• Developer efficiency: Boost execution speed for developers producing and consuming the supergraph.

Assess your graph

The SAF assessment breaks down each pillar into questions on topics like schema design, organizational structure, and supergraph onboarding. SAF automatically saves your progress, allowing you to return and finish the assessment later.

Report generation

To generate a report, finalize your responses by clicking Save on each one. Once you've saved a response for topic, you can click Mark complete on your assessment in the left sidebar. A download icon appears so you can download and share the report with others in your organization.

The assessment rates your supergraph on the five pillars and offers an improvement plan with best practices to implement. Each best practice has an impact and level of effort you can use to prioritize your next steps.

Impact

The impact ratings are:

• Low: Small gains at the individual or team level
• Medium: Changes that can provide noticeable value for individuals or teams
• High: Significant value gained across many teams and the organization

Effort levels

The effort levels use "T-shirt size" estimates ranging from extra small (XS) to extra large (XL).

Next steps

Typically, you should identify best practices with low effort (XS or S) and sort them by highest impact. Actioning best practices in this order provides the highest impact on your supergraph as quickly as possible. Discuss higher-effort (M to XL) items with your team to determine priorities and timelines. If you would like to discuss your assessment in detail, contact your Apollo customer success team.

After you've made progress on your improvement plan, you can retake the assessment to assess your supergraph's new state.