Rajiv Iyer

Sr. Software Engineer

Royal Bank of Canada•  August 2023 - February 2025•  Toronto

Architected and deployed secure workloads using Trusted Execution Environments (TEEs) - Intel SGX and AMD SEV, protecting sensitive data across100% of runtime operations by following Confidential Computing principles, for multi-party data-collaboration with guaranteed confidentiality. Execute security-critical operations inside TEE, while executing other operations outside TEE - to achieve a balance of security & performance. Designed & implemented “CMK Rotation” & “CEK Rotation” for “Always Encrypted Azure SQL” via code, with iterative encrypt / decrypt operations,Secure Key Release from Azure Key Vault (Managed HSM) - inside the TEE - guaranteeing key-security, with an eye on fast-execution. Designed & Implemented key aspects of “Machine Learning based Confidential Inferencing” - uploading ML Models, extracting Model Metadata, running Inference Workloads with large datasets - with the ability for “concurrent high-performance collaborative confidential inferencing”. Optimized run-time performance using code-profiling tools, resolve tricky bugs affecting concurrent execution of 1000s of Jobs in Production Workload conditions, by analyzing log-files via Splunk, reverse-engineering the code. Also improved system observability via Splunk Dashboards. Extensive work in Python, Java, C/C++, Docker / Kubernetes / Ansible / Azure / AWS, Microservices, APIs & ETL, Splunk.

Sr. Software Consultant

IBM Canada•  November 2019 - December 2021•  Toronto

“HPAC” project under “Spectrum Computing” (Data & AI): design, implementation, algorithms, enhancements, optimizations for high-performance computing “pod scheduler” for compute-intensive workloads & integrating the daemons with Container/Orchestration frameworks. Programming in C/C++, Java & Golang for Scheduler components (involving scheduling, ordering & prioritization, preemptions, evictions, etc. - applicable at job-level & pod-level), with DevOps & CI / CD pipelines, and daily exposure to Container Tech, such as: Docker, Kubernetes, OpenShift, Ansible. Designed, Implemented & fine-tuned advanced algorithms tackling complex competing considerations for job/pod scheduling stability vs preemptions, containerizing daemons into pods, resulting in substantial improvements for stabilizing scheduling & balance preemptions, while optimizing run-time using code-profiling tools (pprof), resulting in improving scheduler efficiency by 25%, reducing preemptions 20%, and improving job throughput by 30%.

Sr. Software Engineer

Capgemini•  October 2018 - July 2019•  Mumbai

2019Performed root cause analysis (RCA) through reverse engineering and examining crash dumps to resolve critical issues & feature enhancements forWindows Device Drivers for Dell EMC VNX storage systems, reducing system downtime by 40%, improving incident resolution time by 30%, andenhancing reliability across 1000+ deployed endpoints. Adding support for SMB/CIFS v2.0 protocol. [C, C++, C#, Java, TFS, Visual Studio].

Software Engineer

KAL ATM Software Pvt. Ltd.•  April 2014 - August 2017•  Mumbai

Platform-level C/C++ code to talk to hardware (card-reader, cash-dispenser, etc.) through CEN/XFS protocols, Application-level C#& Java code to build applications specific to various banks/clients, and Front-End through HTML/JS/NodeJS, across 5000+ ATMs & banking terminals. Wrote code which ran across a range of ATM Hardware Vendors & OEMs, while being optimized for fast and secure execution. Led on-site client consulting engagements across 15+ global client locations (Dubai, Turkey, Italy, Tanzania, UK, Norway, among other places).Mentored 20+ new engineers, streamlined processes, and automated workflows, resulting in faster and smoother on-boarding of clients. Facilitated technical seminars and workshops attended by 100+ engineers, promoting engineering best practices and collaborative brainstorming.

Software Engineer

Factset Research Systems•  August 2011 - April 2013•  Connecticut, USA

2013Ensured system readiness and managed 10-minute on-call responses for critical deployments on Unix / OpenVMS, using Perl, Ruby, Python, DCL, Bash, and C/C++, supporting 24/7 operations across multiple servers, across two critical production data-centers in New Jersey & Virginia. Reduced production crashes by 50% through staggered installs & selective code promotions. Refactored 30K+ lines of legacy code. Mentored 20+ engineers during critical incidents and led internal sessions on MapReduce to drive team-wide technical growth.