video1.0<iframe src="https://www.loom.com/embed/1c0a5d7a436c44ddbc45d49f983dd3b2" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/1c0a5d7a436c44ddbc45d49f983dd3b2-836062b9658d4d93.gif2364.188The meeting focused on the use of REST API keys for sensitive data detection, AI classification features, future automation actions, and performance impacts of real-time scanning. Nirav Ganju Cass explained the customization options available with API keys and discussed potential AI integration with Forge LLM to enhance data security. Future plans include expanding automation capabilities through web requests, and it was confirmed that real-time scanning does not significantly impact performance. Next steps involve exploring these features further and addressing customer inquiries. ### Rest api keys for sensitive data detection 0:00 - Hana Packford introduced the topic of using API keys with REST API functionality for sensitive data detection. - Nirav Ganju Cass explained that API keys can be generated within compliance and configured by admins to limit their scope to specific spaces or document types. - The API allows for viewing, managing, and redacting sensitive data based on permissions. - Use cases include flagging generic keys for third-party software checks and setting up exclusion rules for pages. - The REST API provides a comprehensive tool for managing sensitive data programmatically. ### Ai classification features and future integration 2:20 - Hana Packford raised a customer question about AI features in classification and their potential integration with sensitive data. - Nirav Ganju Cass discussed the current use of AI classification with AWS Bedrock LLM and plans to explore Forge LLM functionality with Atlassian. - The goal is to enhance data security by processing data within Forge, eliminating egress from the instance. - Future AI integration aims to provide more flexible and context-aware detection types beyond regular expressions. ### Future automation actions 4:14 - Hana Packford inquired about future plans for automation actions related to compliance. - Nirav Ganju Cass mentioned the intention to move away from explicit actions towards more flexible automation rules using web requests. - This approach allows admins to configure integrations with interconnected systems and leverage teamwork graphs. - The combination of API and automation tools aims to provide a comprehensive suite for programmatic compliance management. ### Performance impact of real-time scanning 6:15 - Hana Packford asked about the performance impact of constant real-time scanning on customer compliance instances. - Nirav Ganju Cass explained that in cloud environments, real-time scanning has no significant performance impact due to Atlassian's infrastructure scaling. - In data center environments, performance impact is minimal, with scans scaling with time rather than processing intensity. - No significant performance issues have been reported by customers or observed in testing.