Self-hosted AI

This is the official documentation of the @forestadmin/agent Node.js agent.

Self-hosted AI

By default, AI features in Forest Admin are processed by Forest Admin servers. To keep your data private, you can route AI requests through your own agent using the addAi method—your data never leaves your infrastructure.

Installation

npm install @forestadmin/ai-proxy

Configuration

import { createAiProvider } from '@forestadmin/ai-proxy';

const agent = createAgent(options)
  .addDataSource(/* ... */)
  .addAi(
    createAiProvider({
      name: 'my-assistant',
      provider: 'openai',
      apiKey: process.env.OPENAI_API_KEY,
      model: 'gpt-4o',
    }),
  );

Custom LLM endpoint

You can point to any OpenAI-compatible API (self-hosted LLM, Azure OpenAI, etc.) using the configuration option:

import { createAiProvider } from '@forestadmin/ai-proxy';

.addAi(
  createAiProvider({
    name: 'my-assistant',
    provider: 'openai',
    apiKey: process.env.LLM_API_KEY,
    model: 'my-model',
    configuration: {
      baseURL: 'https://my-llm.example.com/v1',
    },
  }),
);

Options

Option	Type	Required	Description
name	string	Yes	Unique identifier for this AI configuration
provider	string	Yes	AI provider ('openai')
model	string	Yes	Model to use (e.g., 'gpt-4o', 'o3-mini')
apiKey	string	No	API key (defaults to OPENAI_API_KEY env var)
configuration	ClientOptions	No	OpenAI client options (baseURL, timeout, defaultHeaders)

All OpenAI client options are supported.

Supported models

Any OpenAI-compatible model with function calling support works. Legacy models without function calling (gpt-4, gpt-3.5-turbo, completion models) are not supported.