July 2025 Summaries

Jesse Seldess, the documentation lead at Pinecone, has restructured the company's documentation approach to prioritize user expectations over internal product hierarchies, enhancing ease of use for new users. By leveraging AI tools like Claude Code as a peer reviewer, Jesse has streamlined content optimization, making documentation a self-serve resource that reduces the need for human interaction and supports business scalability. He emphasizes hiring technical writers who can think from a user's perspective and collaborate across teams to produce content that meets real developer needs. Jesse's method includes identifying user goals through personal testing, competitor analysis, and stakeholder collaboration, focusing on reducing friction and increasing clarity for users. His approach highlights the importance of organizing information around user context and using AI as an aid in refining documentation, rather than generating it entirely.

This guide outlines the process of creating and hosting llms.txt files, which are plain Markdown files that provide a concise summary of a website's most important content to aid AI tools in indexing. It emphasizes the benefits of using Mintlify's free generator tool to create starter templates and discusses hosting requirements, ensuring the files are placed at the root of the website for easy access by large language models (LLMs). The llms.txt format includes a title, blockquote summary, and organized sections with links to key content areas, making it particularly useful for product documentation to ensure AI-generated answers are accurate and contextual. The guide also highlights that Mintlify automatically generates and hosts these files, allowing companies to make their content AI-ready seamlessly.

Mintlify has acquired Trieve, a provider of retrieval-augmented generation (RAG) infrastructure, to enhance the AI search capabilities in its documentation platform. This acquisition aims to combine Trieve's high-performance retrieval technology, originally designed by developers Nick and Denzell to overcome limitations in existing vector search tools, with Mintlify's AI-native platform to deliver faster and more accurate AI-powered answers. Trieve's technology will support Mintlify's vision of transforming documentation from static pages into a conversational system capable of answering user queries in real time, thereby improving developer experience and meeting the demands of AI-native environments. The Trieve team will join Mintlify to lead infrastructure efforts, and the open-source project will continue to be supported.