Inferable

Inferable is an innovative, open-source platform designed to streamline the creation of large language model (LLM) driven automations. It empowers developers by offering a seamless way to build robust, agentic workflows that scale. The platform features a developer-first architecture with a focus on production-readiness, facilitating efficient and reliable AI-driven tasks.

Key Features

• Managed Agent Runtime: Utilize a ReAct-like agent runtime, which is bolstered by user-defined functions for robust automation.
• Durable Tool Calling: Ensure resilience with features that recover from failures, manage load balancing, and cache results for optimal performance.
• Zero Network Configuration: Eliminate the need for inbound network configurations, simplifying deployment and use.
• Multi-Language Support: Benefit from native Software Development Kits (SDKs) available for languages like TypeScript, Go, and .NET, among others.
• Fully Open Source: The entire platform is released under the MIT license, ensuring it is fully open source and self-hostable.

How It Works

Inferable allows automations to be powered directly by user code, enabling high levels of customization and integration. Here’s a quick guide to using Inferable:

1. Define Functions: Write functions that will be utilized within your automation.

   async function readWebPage({ url }: { url: string }) {
     const browser = await chromium.launch();
     const page = await browser.newPage();
     await page.goto(url);
     return await page.content();
   }
   

2. Register Functions: Integrate your functions with the platform.

   inferable.default.register({
     name: "readWebPage",
     func: readWebPage,
     input: z.object({
       url: z.string(),
     }),
   });
   

3. Execute a Run: Initiate a process to utilize your registered functions.

   inferable.run({
     initialPrompt: `
       Produce me a list of all the open source projects
       at https://news.ycombinator.com/show, with their
       github url, programming language and fork count.
     `,
     resultSchema: z.array(
       z.object({
         url: z.string().regex(/^https:\/\/github\.com\/.*$/),
         language: z.string(),
         forkCount: z.number(),
       })
     ),
   });
   

Development Infrastructure

Inferable's repository encompasses both the core services and SDKs for a comprehensive development experience:

• Core Services
  • /control-plane for central management.
  • /app for the user-friendly web console.
  • /cli for command-line interactions.
• SDKs included for
  • Node.js/TypeScript
  • Go
  • .NET
  • React
• Bootstrap Templates: Get started quickly with templates for various languages.

Resources

• Quick Start Guide: Comprehensive steps for beginning with Inferable.
• Self Hosting Guide: Instructions for self-hosting.
• Documentation: Full documentation to leverage the full capabilities of Inferable.

Inferable represents a powerful tool for developers looking to integrate intelligent, scalable automation within their applications, providing a straightforward and efficient route to deploying AI-driven functionalities.