Arun Mahendran
Advisory Engineer @ Lenovo
Bengaluru, India
Actions
I am a seasoned System Software Engineer and Data Science & Machine Learning Architect with over 18 years of experience in building software systems using Python, C/C++, and advanced scripting technologies. My expertise spans cloud computing, machine learning, embedded systems, and the design of high-availability and disaster-recovery solutions for enterprise SAN and server storage platforms.
In recent years, I have specialized in data science and data engineering on the AWS cloud, delivering scalable, production-grade analytics and AI solutions. I bring a strong blend of low-level systems engineering and modern AI/ML architecture, enabling me to solve complex, real-world problems across the full technology stack.
I am deeply passionate about data center sustainability and leveraging AIOps to drive carbon reduction, energy efficiency, and intelligent infrastructure optimization at scale.
Links
Badges
Topics
Introduction of the EFI Development Kit (EDK II) for Server Platforms
The EFI Development Kit (EDK II) is a powerful and flexible framework is a modern, open-source, cross-platform firmware development environment for the Unified Extensible Firmware Interface (UEFI) essential for modern server platforms.
This presentation will provide an overview of EDK II, its key features, and how it can be used to develop robust and secure firmware for server systems. We will cover the basics of EDK II, including its architecture, development process, and best practices.
EDK II is a modular framework that supports the development of UEFI firmware, providing a standardized and extensible environment. It includes a rich set of libraries, drivers, and tools for UEFI-compliant firmware, ensuring compatibility and security. The development process involves setting up the environment, building firmware with configuration options, and testing for reliability and performance. Best practices include modular design for maintainability, implementing security features like secure boot and updates, and maintaining thorough documentation.
This session was presented in OCP Global Summit 2025 in San Jose. You may find the session schedule here https://ossindia2025.sched.com/.
Sustainable Software Design: Addressing Bloat and Frameworks with Resource-Proportional Design
Sustainable (carbon-aware)software design is crucial for reducing the environmental impact of software. Challenges like software bloat & the use of frameworks can hinder resource optimization & energy efficiency. Resource-proportional design addresses these issues by tailoring resource usage to the application's specific needs, thereby reducing environmental impact.
Sharing & reusing processed data and models minimizes redundant computations & storage, further lowering carbon production. The ACM paper "From Roots to Fruits: Exploring Lineage for Dataset Recommendations" https://dl.acm.org/doi/10.1145/3600046.3600053 promotes efficient data sharing through dataset lineage, optimizing resource usage and reducing energy consumption in processing pipelines. In this session, we will cover sustainable software design, addressing software bloat & framework inefficiencies with resource-proportional design & discuss data reuse to minimize redundant computations and reduce carbon production.
This session was presented in OCP Global Summit 2025 in San Jose. You may find the session schedule here https://2025ocpglobal.fnvirtual.app/a/schedule/.
AI assist Unit and Integration test framework for OpenBMC
The complexity involved in simulating and injecting hardware failures e.g PSU, DIMM, CPU errors, HDD failures, poses significant challenges for OpenBMC validation. Limited availability of faulty hardware & the high cost of physical test setups, including specific hardware, worsen these issues. Additionally, module-based code development increases the risk of missing existing or domain-specific functional scenarios during new code development. To address these challenges, we propose a unit & integration test framework that reduces the need for actual hardware. This framework uses stubs to simulate firmware code, converts C++ code to Python modules using the SWIG framework & writes function test cases for each module by simulating hardware faults. Beyond traditional testing, we will use AI for code coverage analysis, calculate Cyclomatic Complexity, perform automated patch analysis, vulnerability classification, and fault prediction using historical data and machine learning models.
This session was presented in OCP Global Summit 2025 in San Jose. You may find the session schedule here https://2025ocpglobal.fnvirtual.app/a/schedule/.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top