Main contributors on the call were Valli Guduru, Chrystal Benson, Jay Stotler, and John Buhrmann but others were present. ________________________________ From: Nrupen Bhavsar, Ph.D. <nrupen.bhavsar@duke.edu> Sent: Friday, March 13, 2026 12:29 PM To: Gilani, Isma <igilani@renci.org>; Krishnamurthy, Ashok Kumar <ashok@renci.org> Cc: mism-external@lists.renci.org <mism-external@lists.renci.org> Subject: Re: Urgent: Azure Environment Configuration Blocker for MDSC Platform Hi Isma, Can you clarify who you spoke with within Duke IT? I can send an email to Cliburn (will cc you both on it). Thank you, Nrupen From: Gilani, Isma <igilani@renci.org> Date: Friday, March 13, 2026 at 12:23 PM To: Nrupen Bhavsar, Ph.D. <nrupen.bhavsar@duke.edu>, Krishnamurthy, Ashok Kumar <ashok@renci.org> Cc: mism-external@lists.renci.org <mism-external@lists.renci.org> Subject: Urgent: Azure Environment Configuration Blocker for MDSC Platform Hi Nrupen and Ashok, I wanted to provide an update from this morning’s meeting with Duke IT regarding the Azure environment required to support MDSC platform development (including Kubernetes clusters and public VMs for iRODS). We reviewed the proposed platform architecture and no concerns were raised. However, Duke IT cannot provision the necessary permissions, resources, or configurations until they understand the sensitivity level of the data that will flow through the system. As a result, development in the Azure environment is currently blocked until this information is clarified. To proceed, Duke IT has asked us to provide the following: * A data flow diagram showing where data originates, how it moves through the system, and what is exposed through APIs * RENCI is actively working on this * The data sensitivity level of information entering or leaving the platform * We will need to get this information from the RPs ASAP. Cliburn may have more info as well. * To our knowledge, RP1 may have sensitive data from private BCR repertoire data. * To our knowledge, RP3 may have sensitive data as well. We are less sure what that data is and where it comes from. * Which Duke HPC systems the project will use (identified vs. de-identified patient data environments) * Confirmation of user onboarding and identity management (we indicated MDSC would manage onboarding and users will not require direct Azure access) * Incident response ownership (I indicated Duke) * Responsibility for maintaining the Kubernetes clusters and VMs in the subscription (initially stated Duke) The immediate next step is to clarify the data sensitivity questions and produce the data flow diagram so Duke IT can determine the appropriate environment configuration. I recommend that we make this the first agenda item in Monday’s meeting so we can resolve the open questions quickly. Afterward, we will need to schedule a follow-up call with Duke IT and Duke Health CP to review the responses. It would be helpful for Nrupen to join that discussion. Best, Isma