Making Alt Text Fast: How Twilio Scaled Docs Accessibility with Automation

Twilio has successfully enhanced the accessibility of its documentation by automating the generation of alt text and image filenames through a combination of scripts and large language model (LLM) prompts. This approach was driven by the need to provide descriptive text alternatives for images on their documentation site, a task that had become cumbersome and inconsistent without automation due to the sheer volume of images involved. By leveraging respected accessibility guidelines and integrating them into a command-line tool and a web application called Docs AI Buddy, Twilio streamlined the process, allowing for efficient batch processing and ensuring human review of alt text prior to publication. The automation has resulted in significant productivity gains, with the LLM-generated descriptions passing human review for 99.96% of images and reducing the time required for processing and reviewing alt text from 15-30 minutes to just a few seconds. This initiative not only bolstered accessibility but also freed up Twilio's content teams and engineers to focus on other high-impact tasks, demonstrating the effectiveness of using AI to manage large-scale documentation challenges.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.