Core Features
This section covers the features most teams touch first when adopting TramAI: how to think about the library, how providers and models are wired, and how structured, tool-enabled, and streaming operations behave in practice.
In This Section
| Guide | Focus |
|---|---|
| Why TramAI | Positioning, design goals, and why the project is built around typed AI operations |
| Why TramAI Workflows | When to move from single calls to workflow-oriented orchestration |
| Providers and Model Routing | Provider registration, model mapping, and routing decisions |
| Structured Output | Schema generation, parsing, validation, and retry behavior |
| Tool Calling | Registering tools, governance enforcement points, and approval suspension |
| Streaming | Incremental response handling and streaming execution patterns |
| Spring Boot Integration | The unified tramai-spring-boot-starter, tramai.profile, and YAML-driven wiring |
When To Start Here
Start here if you are learning TramAI’s core mental model or setting up the first production-style AI integration in a JVM service.