Apollo has pivoted sharply toward AI integration, positioning GraphQL as the foundation for intelligent agents. The blog shipped foundational pieces—Apollo MCP Server, Apollo Skills, and type-safe patterns for connecting LLMs to APIs—then spent the quarter exploring practical applications: orchestrating multiple APIs for agents, making existing schemas AI-ready, and building secure agent stacks. A parallel track addressed operational maturity, with GraphOS Platform API enhancements and a deep dive on telemetry cardinality in the Router, while Apollo Client 4.1 added streaming capabilities that serve both traditional and agentic use cases.
March centered on AI integration patterns. The blog shipped four posts on building AI agents with GraphQL—from schema-driven agents to API orchestration for LLM backends—alongside coverage of the new Apollo Kotlin normalized cache and MCP (Model Context Protocol) server implementations for connecting GraphQL APIs to AI systems.