Case Study: Taming API Sprawl with a Federated GraphQL Gateway

After Phase 1 (independent monoliths) and Phase 2 (event sourcing for shared facts), our frontends reached the point where a single app needed data from multiple domains. Service‑specific REST endpoints no longer fit. Frontend code was orchestrating multiple calls and model translations.

We needed a gateway. After research and prototyping, we chose a federated GraphQL architecture with ChilliCream Hot Chocolate Fusion because our services already fit the idea of nodes in a larger domain graph.

The problem: the tangle of microservices

As our domain expanded, we faced a familiar set of issues:

• API sprawl. Each service had its own REST endpoint and model.
• Coupling risks. Changes in one service leaked into consumers.
• Duplicated logic. Clients stitched data manually.

The solution: a federated gateway with Hot Chocolate Fusion

Instead of a simple proxy, the gateway composes service schemas into a unified super‑schema, plans queries, fans out to the right services in parallel, then stitches the results.

1. Each microservice is a GraphQL server. Self‑contained, owning its domain.
2. Gateway composes schemas. It knows which service resolves which fields.
3. Clients make one query. Ask for exactly what they need across domains.

A key design choice: keep complex joins out of individual services. By exposing the right identifiers and boundaries, the client and gateway can compose cross‑domain views without creating backend dependencies.

Product impact for Auxin

• One cohesive API for all apps. Faster feature work and a simpler frontend mental model.
• Preserved team autonomy. Services stay independently deployable and evolvable.
• Strong DX. Strongly typed, self‑documenting contracts.

How this fits the bigger picture

• Phase 1 gave us fast, isolated delivery: /writing/independent-monoliths-platform
• Phase 2 gave us shared facts without coupling: /writing/event-sourcing-at-auxin
• Phase 3 unified the client surface with a graph: this post

This is where Auxin became what it is now: a place where future‑oriented growers analyze diverse greenhouse data in a single view, and where specialized apps can go deep on the same facts, all on one secure platform.