Transcript
by deploying intelligent assistants that answer employee questions, streamline support, and enhance operational efficiency. But there's a catch. The knowledge that powers these models is scattered across legacy file servers, buried in HR documents, contracts, and archived project folders. Some of this data is perfect for AI. Some of it is highly sensitive and absolutely cannot be exposed. Compliance officers need to confirm that only approved data is used, while data scientists require fast and reliable access to high-quality datasets to build and test AI models. Too often, these teams are stuck trying to balance speed with security and innovation with governance. In this demo, you'll see how Data Rooms addresses these challenges by helping compliance officers identify and control sensitive data before it reaches any AI pipeline, giving data scientists access to curated, compliant datasets without waiting for manual approvals or risking policy violations, streamlining the entire workflow from risk analysis to AI-ready export for transparency and auditability. We start with the compliance officer who has already run Commvault risk analysis across their company's file servers. Sensitive data, such as personal identifiers in HR files or confidential financial information, is flagged. Low-risk content, like process guides and internal FAQs, is identified as safe for AI use. This automated process helps provide compliance officers with assurances that sensitive data is detected and controlled, reducing the need for manual review and the risk of accidental exposure. With a few clicks, the compliance officer creates a new Data Room, including only the approved low-risk files that can be safely used to train a secure internal knowledge assistant. Access is granted to the data science team, who will use this curated data to prepare and test the Retrieval Augmented Generation model, allowing only compliant data to be used for AI training. The new Data Room appears in Data Studio. The software is now ready to be used as a central workspace for data scientists. Each Data Room is clearly labeled with its name, creation date, and owner. This organized view helps prevent the need to chase down ad hoc file requests or wait for approvals. Data scientists can easily see which data sets are ready and approved for use and begin working within the governed environment. Opening the Data Room, the data scientists can browse a curated set of low-risk files, such as internal policies, IT runbooks, and departmental handbooks. These documents are ideal for powering a RAG model, enabling employees to ask questions like, what's our onboarding process? Or, how do we restore a server from backup? The data scientist can search, filter, and review the data set, knowing that every file has already undergone compliance review. This means they can focus on building and testing models, not second-guessing data quality or compliance. Once the data set is ready, the data scientist initiates an export request to the secure AI Data Lake. They select the AI Data Lake as the export destination, a secure S3-compatible storage location where files will be prepared for vectorization and indexing in the RAG pipeline. The export request automatically includes all approved files and permissions, preventing the need for manual configuration or path selection. The compliance officer receives a notification to review this export request, retaining final approval over any data leaving the protected environment. They open it, confirm that only low-risk folders are included, check the destination details, and approve it. This step helps to maintain data security and speed for the AI project. Back in Data Studio, the data scientist can monitor the export's progress, viewing status updates, transfer logs, and complete details for each export. Once complete, the team has a clean, governed data set ready for ingestion into the RAG knowledge base. Each stage, from risk detection to export, is traceable, giving compliance and AI teams confidence that the data foundation is secure and reliable. In just a few minutes, we've moved from risk detection in raw file data to delivering a compliant, AI-ready data set, while avoiding bottlenecks or compliance headaches. What's powerful here is how seamlessly the compliance officer and data scientist collaborate, each focused on their part of the process, yet remaining connected through a single, governed workflow. What sets Data Rooms apart is its ability to provide compliance officers with complete control, defining what's safe and confirming policy alignment at every step. It also allows data scientists to access curated data quickly, accelerating AI and analytics projects while maintaining strict governance. Additionally, every action taken within Data Rooms is auditable and traceable, giving both teams confidence that the data foundation is safe, reliable, and ready for the future of AI.
