autoflow: what it is, what problem it solves & why it's gaining traction

What it solves

AutoFlow provides a way to build a knowledge base using Graph RAG (Knowledge Graph Retrieval-Augmented Generation), allowing users to create conversational search experiences for their websites or documentation.

How it works

It combines several technologies to process and retrieve information:

• Data Collection: It uses a built-in website crawler to scrape sitemap URLs from official documentation sites.
• Storage: It uses TiDB Vector to store chat history, vectors, JSON, and analytics.
• Orchestration: It leverages LlamaIndex as the RAG framework and DSPy for programming foundation models.
• Interface: It provides a Perplexity-style conversational search page and an embeddable JavaScript snippet for adding a search widget to existing websites.

Who it’s for

Developers and product owners who want to implement an AI-powered conversational search or a knowledge base based on their own documentation.

Highlights

• Graph RAG: Utilizes knowledge graphs for improved retrieval.
• Built-in Crawler: Effortlessly scrapes documentation sites via sitemaps.
• Embeddable Widget: Easy integration into websites via a JavaScript snippet.
• Modern Tech Stack: Built with Next.js, Tailwind CSS, and shadcn/ui.