Divyanshu Madan Email & Phone Number

Bengaluru, Ka, In

Bengaluru, Karnataka, India

• Spearheaded the architectural design of scalability workloads, devising a solution featuring CRUD operations scheduling via Docker daemons on a bootstrap container. Achieved 95% reduction in image size by optimizing containers with Docker Slim.• Conducted extensive research on various Kubernetes (K8s) distributions, including kind, k3s, and k3d, as well as virtual cluster solutions like vCluster and mocking tools such as Kwok, to identify optimal solutions for scalability workloads.• Led the conceptualization and implementation of Gatekeeper workflows, covering image registry, security, and mutation policies. Developed validation pods and deployment configurations to ensure system compliance in Tanzu environments.

Bengaluru, Karnataka, India

• Led the design and development of chaos engineering strategies, leveraging LitmusChaos to craft custom experiments tailored for Tanzu’s K8s Custom Resource Definitions (CRDs), enhancing fault tolerance and resilience within the Tanzu Application Platform (TAP).• Implemented top-level keys in user inputs to optimize configuration sharing across diverse components, enhancing user experience during installation processes.• Crafted a user-friendly CLI installer using Cobra commander in Go, distributed for Linux, macOS, and Windows environments, that guide users through simple and advanced configuration options, facilitating effortless execution during installations.• Developed a Go-based interface to streamline stack promotions across development, staging, and production environments, simplifying the transition between different deployment stages.

Bengaluru, Karnataka, India

• Led the conceptualization and implementation of resiliency workflows for various supply chain and multi-cluster specifications, featuring benchmarking capabilities. Deployed on Kubernetes clusters across Amazon (EKS), Azure (AKS), Google (GKE), and VMware (TKG) environments.• Crafted Go modules for various functions of TAP’s developer workflows, utilising the Kubernetes API. Integrated these modules into frameworks focused on resiliency, performance, scalability, and longevity.• Engineered automated release packaging, incorporating image repository bundles for K8s CRDs, Tanzu CLI, and associated plugins, leveraging Go and Carvel tooling.

• Engineered a silent CLI deployment of vCenter Server Appliance on ESXi hosts, integrating pre-deployment and post-deployment services for hosts, datastores, NTP servers, and licenses via the vSphere API in Python. This eliminated manual work during deployments, resulting in task efficiency improvements ranging from 25% to 99%.• Implemented the solution across three distinct workload deployments, slashing deployment times by ≈ 5 hours (28% improvement), ≈ 10 hours (41% improvement), and ≈ 12 hours (99% improvement) respectively.• Developed a comprehensive health monitoring and alerting system for in-production ESXi hosts utilizing the vSphere API and system log analysis. This system monitors CPU, memory, storage, and critical parameters like kernel panic and memory leaks, and proactively detects issues for critical servers, ensuring smooth deployments and minimizing downtime.

• Developed Python interface for Point Cloud Library (PCL), enhancing usability while preserving C++computational speed.• Engineered utilities for static analysis of native language types using the LibClang API, enabling in-depth AST introspection and CMake artifact analysis.• Designed and implemented a pybind11-based system to create a Python interface, leveraging semanticrepresentation and build system metadata for binding to native C++ API.

• Conducted an exhaustive literature review on image quality assessment (IQA) methodologies, focusing on deep learning techniques. Examined subjective and objective assessment methods for evaluating perceptual quality and technical precision.• Carried out extensive experiments to evaluate Google’s Neural Image Assessment (NIMA) across various convolutional neural network (CNN) architectures and training datasets, assessing its performance on a custom testing dataset. Implemented a no-reference IQA workflow in Python with TensorFlow, optimizing the MobileNet architecture through fine-tuning with the TID2013 dataset.

Manipal, India

• Spearheaded architecture and development for Simultaneous Localization and Mapping (SLAM) subsystem in Formula-style autonomous driving pipeline.• Conducted extensive research on SLAM techniques, including Kalman filters, particle filters, FastSLAM2, and ORBSLAM2 algorithms, to identify optimal solutions.• Designed and implemented robust solution utilizing Error State Extended Kalman Filter, ensuring high precision in localization and mapping tasks.• Conducted comprehensive simulations using Robot Operating System (ROS) and CARLA, validating system performance and reliability across various real-world scenarios and environmental conditions.

Pune, Maharashtra, India

• Conceptualized and engineered off-board diagnostics for Anti-lock Braking System (ABS) and Electronic Stability Program (ESP®) modules using Unified Diagnostic Services (UDS) communication protocol.• Implemented diagnostic routines by crafting hexadecimal message frames from service IDs and parameters for execution on the Electronic Control Unit (ECU) over the Controller Area Network (CAN) bus. Analyzed and interpreted controller responses to compile comprehensive data logs detailing ECU environment and operating conditions.• Developed a C-based keygen for facilitating secure access to critical ECU services through a seed-key exchange protocol.