Skip to content

Runtime Tutorials

Progressive learning path for building LLM applications with Runtime.

Learning Path

Follow these tutorials in sequence to build expertise:

1. Getting Started

First Pipeline - Build your first LLM application (15 min)

2. Conversations

Multi-Turn Conversations - Add conversation state and memory (20 min)

3. Tools & Integration

MCP Integration - Connect external tools via MCP (30 min)

4. Safety & Validation

Validation & Guardrails - Add content safety checks (20 min)

5. Production Ready

Production Deployment - Build production-ready apps (40 min)

6. Advanced Topics

Advanced Patterns - Optimization and advanced techniques (30 min)

What You’ll Learn

Tutorial 1: First Pipeline

Create and configure pipelines
Connect to LLM providers
Execute basic requests
Handle responses and costs

Tutorial 2: Multi-Turn Conversations

Manage conversation state
Implement Redis persistence
Handle context windows
Build chatbot interfaces

Tutorial 3: MCP Integration

Set up MCP servers
Register external tools
Handle tool calls
Build tool-enabled agents

Tutorial 4: Validation & Guardrails

Implement content filters
Add custom validators
Handle validation errors
Build safe applications

Tutorial 5: Production Deployment

Error handling strategies
Monitoring and logging
Cost optimization
Scalability patterns

Tutorial 6: Advanced Patterns

Multi-provider fallback
Streaming optimization
Custom middleware
Performance tuning

Prerequisites

Go 1.21 or higher
Basic Go programming knowledge
LLM API keys (OpenAI, Claude, or Gemini)
Terminal/command line familiarity

Time Estimate

Quick Start: Tutorial 1 only (15 min)
Core Skills: Tutorials 1-3 (65 min)
Production Ready: Tutorials 1-5 (125 min)
Complete Path: All tutorials (155 min)

Next Steps

Start with First Pipeline to build your first LLM application.

See Also

Runtime Reference - Complete API documentation
Runtime How-To - Task-focused guides
Runtime Explanation - Architecture concepts