Building AI Applications with PHP

Building AI Applications with PHP

1. A Practical Guide Using PapiAI, Laravel, and Symfony

Part I: Foundations

Chapter 1: The AI-Native PHP Application

1. What You’ll Build
2. Who This Book Is For
3. What You Need
4. The Anatomy of an AI Application
5. Installing PapiAI

Chapter 2: Your First Agent

1. The Provider
2. The Agent
3. The Response
4. Streaming
5. Swapping Providers
6. Failover

Chapter 3: Tools — Giving the AI Hands

1. How Tool Calling Works
2. Creating Tools with Closures
3. Creating Tools with PHP Attributes
4. Tool Design Principles
5. The Agentic Loop and maxTurns

Chapter 4: Conversations and Memory

1. The Conversation Object
2. Messages
3. Why Context Management Matters
4. Context Window Strategies
5. Persisting Conversations

Chapter 5: Structured Output

1. The Schema System
2. Schema Types
3. Constraints and Modifiers
4. Validation

Chapter 6: Retrieval-Augmented Generation (RAG)

1. How RAG Works
2. Embeddings with PapiAI
3. The Vector Store
4. Building a RAG Pipeline
5. RAG as a Tool
6. Tuning RAG Quality
7. Production Considerations

Chapter 7: Middleware

1. The Middleware Interface
2. Built-in Middleware
3. Stacking Middleware
4. Custom Middleware

Part II: Building the Laravel Project

Chapter 8: Project Setup — The Support Agent

1. Scaffolding
2. Configuration
3. Database Setup
4. The System Prompt

Chapter 9: Building the Tools

1. OrderLookup
2. KnowledgeBase
3. InventoryCheck
4. Registering Tools in the Service Provider

Chapter 10: The Chat Controller and Streaming

1. The Controller
2. Routes
3. The Frontend
4. Testing the Flow

Chapter 11: Security

1. Prompt Injection
2. Tool Security
3. API Key Management
4. Rate Limiting
5. Cost Controls

Part III: Building the Symfony Project

Chapter 12: Project Setup — The Content Pipeline

1. Scaffolding
2. Bundle Configuration
3. Database Schema
4. The Article Schema

Chapter 13: The Content Generation Pipeline

1. The Pipeline Orchestrator
2. Async Processing with Messenger
3. The Controller
4. Running the Worker

Chapter 14: Performance

1. Streaming
2. Caching
3. Model Selection
4. Async Processing
5. Token Optimization
6. Connection Reuse

Chapter 15: Architecture Patterns

1. The Agent as a Service
2. Separate Concerns
3. Prompt Management
4. The Router Pattern
5. Observability

Chapter 16: Testing AI Applications

1. Testing Tools
2. Testing with Mock Providers
3. Testing Schemas
4. Integration Tests

Chapter 17: Deployment and Production Concerns

1. Environment Configuration
2. Health Checks
3. Monitoring
4. Timeouts
5. Scaling

Appendix A: Provider Reference

1. Provider Capabilities Matrix
2. Provider Constructor Parameters

Appendix B: Schema Reference

1. Types
2. Constraints
3. Validation Methods

Appendix C: Middleware Reference

1. RetryMiddleware
2. RateLimitMiddleware
3. CacheMiddleware
4. LoggingMiddleware
5. Custom Middleware