Guides
TramAI guides are organized into five groups. Whether you are getting started with your first typed AI service or hardening a production deployment, there is a guide for that.
Core Features
Foundational concepts and essential features every TramAI user should know.
| Guide | Description |
|---|---|
| Why TramAI? | Philosophy, positioning, and how TramAI compares to LangChain4j and Spring AI |
| Providers & Routing | Configure provider modules, model routing, and fallback strategies |
| Structured Output | JSON schema generation, parsing, validation, and retry feedback loop |
| Tool Calling | Register and invoke tools within AI operations |
| Streaming | Streaming responses from typed service interfaces |
| Spring Boot | Auto-configuration, property binding, and @AiTool bean discovery |
Testing & Observability
Ensure correctness and visibility of AI operations in any environment.
| Guide | Description |
|---|---|
| Testing | Mock providers, simulated failures, operation recording, and fluent assertions |
| Observability | OpenTelemetry traces and metrics for every AI call |
| CI Integration | Deterministic testing in CI pipelines without network access |
Production
Run TramAI in production with confidence — deployment, hardening, and operations.
| Guide | Description |
|---|---|
| Production Hardening | Resilience, retry, circuit-breaking, caching, and token budgeting |
| Server Mode | HTTP API surface for TramAI workflows |
| Platform Operations | Consolidated deployment and operations layer |
| Scheduling | Recurring and cron-triggered workflow execution |
| Native Image | GraalVM native-image compilation for fast startup and low memory |
| Troubleshooting | Common issues, diagnostics, and solutions |
Extending TramAI
Build custom integrations, interceptors, observers, and MCP tools.
| Guide | Description |
|---|---|
| Custom Providers | Create provider modules for custom AI backends |
| Custom Interceptors | Add cross-cutting behavior with the interceptor SPI |
| Custom Observers | Implement custom operation observers for monitoring |
| MCP Integration | Expose TramAI workflows as MCP tools over stdio or SSE |
Deep Dives
Architecture, design decisions, and internals for advanced users and contributors.
| Guide | Description |
|---|---|
| Architecture Overview | High-level architecture and module layering |
| Design | Design principles and key abstractions |
| Module Design | How modules are structured and composed |
| Orchestration | Typed workflow orchestration with checkpoint/resume |
| Orchestration Persistence | Persistence layer for checkpoint and memory data |
| Orchestrator Vision | Future direction of the orchestration system |
Quick Links
- Get Started: Quickstart | Tutorial
- Reference: Annotations | Configuration | API Stability
- Security: Sovereign Mode | DLP | Evidence Packs
- GitHub Engineering Artifacts: Engineering Folder