Naina K. Email & Phone Number

United States

United States

United States

- Built Performance Regression Monitoring Pipeline to highlight performance inconsistencies in high performance large-scale distributed Remote Direct Memory Access (RDMA) Networks used for AI training.- GPU-based RDMA networks at Meta are high-performance large-scale distributed systems - at par with top 10 Supercomputers across the world.- Expanded distributed performance benchmark tests to stress high-performance large scale RDMA Networks.- Developed framework to schedule recurring performance benchmark jobs and AI training communication primitives - All-reduce and All-to-all across various data centers.- Designed analysis pipeline for monitoring regressions due to planned/ unplanned maintenance and failures. Generate alerts for informing responsible teams during any observed performance decline.- Framework identified 50% regression in subset of production network designs due to a routing bug.- Helped Network Infra org to define service level performance consistency guarantees offered by network.

United States

Bengaluru, Karnataka, India

Team: Bradbury Infra- Part of the backend team that handled the creation and generation of manageable update package [that contained drivers/firmware/ applications data for installation] adhering to Dell standards.- Effectively owned and managed the Dell Update Package Framework, an installation framework that packages the Drivers, Firmware, and Dell Applications for remote software updates on Dell Endpoints. The framework provides UI/ CLI services to install/ update the software. - Delivered critical feature enhancements and bug fix releases. - Worked on various security enhancements for the project.- Worked on development and management of Compliance Verification Tool (CVT) for automated schema check and software installation verification on endpoints. CVT is a part of the Dell Dominion project, a windows desktop application that manages the creation and verification of manageable update packages.- Collaborated in developing and releasing the Command Line Interface for Dell Dominion that automated the package creation and compliance verification process for our customers reducing their manual efforts drastically. Released the feature within 30% less timeframe than the original engineering effort estimate.Team: Dell Data Protection- C++/C# developer for the client team of Dell Data Protection for the Windows platform, an encryption product that protects data and detects data security risks on desktops, laptops, and external media. - Worked on efficiently onboarding the product from the US team to India for sustenance.- Developed feature enhancements for encryption, authentication, policy enforcement, vault recovery, and inventory uploading. - Handled setting up of debugging environment that helped in reducing the debugging efforts for the team.- Supported the first two major releases with critical enhancements, bug fixes, and security fixes.

Bengaluru, Karnataka, India

Team: Dell Data Guardian• Part of the Dell Data Guardian's Android team working on protecting in-motion and in-use data using encryption and Digital Rights Management. • Worked on a Proof of Concept (POC) to identify the sensitive files on Android Phone and suggest the user to protect them using Data Guardian App. • Used TensorFlow Lite and Keras framework to integrate Machine Learning with the Android application. • Analyzed and Identified efficient Machine Learning Model (Artificial Neural Networks) for the App.• Generated vast amount of data for training and validation sets consisting of simulated sensitive data.• Proved feature's viability with prediction accuracy of 75-80%.

Master Of Science - Ms, Information Security

Bachelor Of Engineering - Be, Computer Science