Workshop

BoxLang AI Bootcamp - Zero to Hero

This workshop is a hands-on, one-day introduction to building **AI-driven applications using BoxLang’s native AI module**. BoxLang provides a unified, fluent API to work with multiple LLM providers, abstracting away the complexity of each provider’s SDK, formats, and workflows. Participants will learn how to integrate, configure, and extend AI functionality (chat, embeddings, tools, pipelines) directly within BoxLang applications. 

Objective

By the end of this bootcamp, participants will be able to:

• Understand BoxLang’s AI module architecture and design principles.
• Configure and use multiple AI providers through a unified API.
• Construct chat interactions, asynchronous AI calls, and custom pipelines.
• Build advanced AI workflows including structured outputs, tools, and memory.
• Integrate AI features into BoxLang applications effectively and efficiently.

Audience

This workshop is designed for:

• Developers with basic experience in JVM languages or BoxLang.
• Engineers looking to integrate AI into applications with minimal boilerplate.
• Technical leads evaluating how AI fits into modular JVM systems.
• Anyone interested in practical AI application development on the JVM.

Outline

• Introduction & Environment Setup
  • BoxLang AI platform overview, bx-ai module installation, provider configuration, API keys, mental model.
• Basic AI Chats
  • Chat requests, options and parameters, message roles, synchronous vs async execution, return formats and responses.
• Multi-Modality & Provider Switching
  • Text vs multi-modal inputs, switching AI providers, provider overrides, portability considerations.
• AI Tools Fundamentals
  • Creating aiTool() functions, structured inputs and outputs, tool safety, integrating tools into basic chats.
• AI Pipelines Basics
  • Pipeline concepts, runnable interfaces, execution contexts, transformers, chaining steps, return formats.
• Agents Fundamentals
  • Agent architecture, memory concepts, runnable agents, tool usage, orchestration patterns.

Prerequisites

• Have BVM or BoxLang Latest installed
• VSCode IDE with BoxLang Module
• API Keys for at least one API Provider:
  • OpenAI - https://platform.openai.com/api-keys
  • Claude - https://console.anthropic.com/login
  • OpenRouter - https://openrouter.ai/
• Docker
  • For running a standalone Ollama server
  • For running vector databases