koan-ai-integration

What problem does it solve?

Integrating AI capabilities like chat, embeddings, and vector search into traditional applications can be complex, requiring separate libraries and data stores. This Skill provides a native, seamless way to embed AI directly into your Koan entities and workflows.

Core Features & Use Cases

• Native Chat Endpoints: Easily expose AI chat functionality through standard controllers, leveraging configured AI providers.
• Entity Embeddings: Store vector embeddings directly on your entities, enabling semantic search and similarity comparisons.
• RAG Workflows: Build Retrieval-Augmented Generation (RAG) systems by combining vector search for relevant documents with AI chat for contextual answers.
• Vector Search: Perform semantic searches on your data, finding related items based on meaning, not just keywords.
• Use Case: Create a product catalog where users can search for "eco-friendly laptops" using natural language, powered by vector embeddings and AI, or build a customer support chatbot that answers questions based on your knowledge base.

Quick Start

To enable AI chat in your application, inject IAi into your controller and use it: public class ChatController : ControllerBase { private readonly IAi _ai; public ChatController(IAi ai) => _ai = ai; [HttpPost] public async Task<IActionResult> Chat([FromBody] ChatRequest request) { var response = await _ai.ChatAsync(new AiChatRequest { Messages = request.Messages }); return Ok(new { message = response.Content }); } }