AILang

English | 中文

This project validates a collaboration model where the program handles stable, deterministic flows, while AI handles business judgment and content generation that are hard to predefine.

What problem does it solve

It fits scenarios like these:

• The goal is clear, but the rules change frequently
• Business results need to be extracted from unstructured inputs such as text, images, and notes
• AI output must continue through the system workflow instead of stopping at the chat interface

Code examples

C#

Console.Write("Enter the image path: ");
var imagePath = Console.ReadLine()?.Trim();

if (string.IsNullOrWhiteSpace(imagePath))
{
    Console.WriteLine("No image path provided. The program has exited.");
    return;
}

imagePath = Path.GetFullPath(imagePath);
if (!File.Exists(imagePath))
{
    Console.WriteLine($"Image not found: {imagePath}");
    return;
}

var result = await _engine.AI<dynamic>(
    "Recognize entities",
    new
    {
        image = AiImageAttachment.FromFile(imagePath)
    }
);

if (result.Exist("cat"))
{
    Console.WriteLine($"Number of cats in the image: {result.CatNumber}");
}
else
{
    Console.WriteLine($"Number of other entities in the image: {result.OtherNumber}");
}

Console.WriteLine($"Image description: {result.Description}");

JS

if (!FILES.length) {
  throw new Error("Please select an image first.");
}

const image = FILES[0];
if (image.kind !== "image") {
  throw new Error("Please select an image file.");
}

const result = await AI(
  "Recognize entities",
  { image }
);

if (result.entities?.includes("cat")) {
  console.log(`Number of cats in the image: ${result.catNumber}`);
} else {
  console.log(`Number of other entities in the image: ${result.otherNumber}`);
}

console.log(`Image description: ${result.description}`);

These examples are meant to highlight that: AI is not limited to natural-language description. It can directly produce results that match how the downstream code will use them, so the program can keep consuming and executing them.