Transforming Documentation Workflows with Amazon Bedrock and VerbaGPT: A Skyflow Case Study

In the fast-paced world of startups, it’s not uncommon for engineering to take precedence over documentation. This can often lead to a mismatch in release cycles, where features are rolled out without corresponding documentation, resulting in increased support calls and unhappy customers. Skyflow, a data privacy vault provider, experienced this challenge as it expanded globally in early 2023.

With a documentation team of only two people and an engineering team of over 60, Skyflow found itself struggling to keep up with the rapid pace of feature releases. The accuracy of their technical content was crucial, as customers trust them with their sensitive data. Despite releasing new features every other week, documentation would take an average of 3 weeks to complete.

To address this issue, Skyflow turned to generative artificial intelligence (AI) to streamline their documentation workflow. They built VerbaGPT, a tool based on Amazon Bedrock, a managed service for foundation models from leading AI startups and Amazon. VerbaGPT uses Retrieval Augmented Generation (RAG) to provide accurate first drafts quickly, reducing the time needed for drafting and reviewing.

By implementing this new AI-powered solution, Skyflow was able to significantly speed up their content creation process. They were able to generate drafts of different content types in minutes instead of days, reducing the time from an average of 3 weeks to as little as 5 days. The security measures provided by Amazon Bedrock aligned with Skyflow’s commitment to data privacy, ensuring sensitive information was protected.

The efficiency and productivity gains achieved by using VerbaGPT have transformed how Skyflow creates technical content. The tool is now integrated with Skyflow LLM Privacy Vault to maintain data privacy standards while using AI for content creation. The journey of building and implementing VerbaGPT showcases the power of AI in enhancing technical content creation.

If you’re facing challenges with keeping up with documentation while scaling your product, consider leveraging AI tools like VerbaGPT to streamline your content creation process. By utilizing technologies like Amazon Bedrock, you can create technical content in days instead of weeks, all while maintaining the highest standards of data privacy and security. Learn more about how Skyflow can help transform your approach to data privacy and contact your AWS representative to see if Skyflow is available in your region.

About the authors:
Manny Silva is the Head of Documentation at Skyflow and the creator of Doc Detective. Jason Westra is a Senior Solutions Architect for AWS AI/ML startups. Both authors are passionate about leveraging technology to improve developer experiences and drive innovation in the industry.