What is our primary use case?
My main use case for Apollo GraphOS is managing and monitoring GraphQL APIs across the development life cycle. I use it to manage our GraphQL schema, validate schema changes before deployment, and monitor API usage and performance. It also helps us collaborate across the teams by providing a centralized view of a GraphQL schema, and it ensures that changes don't introduce breaking issues for existing clients.
For example, when we introduce a new field or modify an existing schema, we use the schema validation and checks to verify that the change won't break existing front-end applications. That gives us more confidence when deploying the new update and reduces the risk of API compatibility issues.
One additional benefit is that Apollo GraphOS provides a centralized platform for our GraphQL API, instead of managing a schema and API changes separately in different services.
How has it helped my organization?
Apollo GraphOS has had a positive impact by improving the way we manage and evolve our GraphQL APIs. Having a centralized schema registry and validation checks has reduced the risk of introducing breaking changes during deployments. While I don't have organization-wide metrics, from my experience, it has reduced the time spent troubleshooting schema-related issues and reviewing API changes. It also made deployments smoother by identifying compatibility issues before they reach production.
I don't have formal organization-wide metrics, but based on my day-to-day experience, I would say our efficiency has improved by around 20% to 25%.
What is most valuable?
The best features I find most valuable are schema registry, schema checks, API observability, and Apollo Federation.
It provides visibility into API usage, performance, and query patterns. This helps us to identify bottlenecks and understand how clients are using GraphQL APIs.
What needs improvement?
Overall, I've had a good experience, but if I had to choose one improvement, it would be improving the onboarding and troubleshooting experience for large GraphQL environments. A better diagnostic tool with more guided information and recommendations would help developers resolve issues more quickly.
I would like to see more customizable dashboards and reporting so that different teams can focus on the metrics that matter most to them.
I rated it six out of ten because while it offers useful capabilities such as schema management, schema validation, and API observability, I feel there are several areas that need improvement. In larger or more complex environments, I've experienced performance limitations and a steep learning curve for some advanced features. I think the troubleshooting experience and documentation could be better.
For how long have I used the solution?
I've been using Apollo GraphOS for about two years.
What do I think about the stability of the solution?
I would say it's reasonably stable on a day-to-day basis. We haven't experienced much outage.
What do I think about the scalability of the solution?
Regarding scalability, I feel it's suitable for many GraphQL workloads. However, in larger or more complex environments, performance will be somewhat reduced.
How are customer service and support?
I feel the customer support is good. I would rate it a ten out of ten. They were always ready to help.
Which solution did I use previously and why did I switch?
Previously, we used manual schema reviews and documentation instead of a new solution, so it was an improvement over our previous approach.
How was the initial setup?
From a developer's perspective, the setup and onboarding into the development environment was straightforward. I felt it was good value for the capabilities that we used.
What about the implementation team?
I was not involved in those discussions.
What was our ROI?
I think we have seen a return on investment. We have improved our development efficiency. I would estimate we saved 10% to 15%, or perhaps 20% of the time previously spent on manually checking schema, schema validation, or API change reviews.
Which other solutions did I evaluate?
We do not have any relationship with the vendor.
What other advice do I have?
A good example was when we needed to update a part of our GraphQL schema to support a new feature. Before deploying the change, we used the schema registry and schema checks to verify whether existing client applications would be affected.
Regarding AI from a governance and security perspective, I think Apollo GraphOS does a good job. Features such as the schema-related checks reduce the risk of breaking changes.
Based on my experience, the platform's core features have been reliable. I don't have sufficient experience with its AI capabilities to rate the accuracy of AI-generated output.
I would recommend that organizations have a clear understanding of their GraphQL architecture and use cases before adopting Apollo GraphOS. They can take advantage of features such as schema management, schema checks, and observability from the beginning, and at the same time, invest time in learning the platform's advanced features. I rated this product six out of ten overall.