Singapore, Singapore
Aspiring to automate society with technology. Being a DD - someone with a deep desire to reach out to the general public, teaching about the advantages of cloud technologies and microservices in order to develop more robust systems. Aspiring to be: An Eventer. Looking to decouple services and to implement EDA. My inspirations in life: TIGER: To be fearless in whatever I do FIRE: Burning with fiery passion CYBER: Striving to learn the best practices from around the globe FIBER: Connecting with anyone across the globe DIVER: Jumping into solving issues VIBER: Being able to communicate with others out of work settings JAJA: y hablo espanol With a background in computer science, I have been involved in cloud migration projects: assessing migration feasibility, architecting the integration of interacting applications on the cloud; with a focus on financial workloads. I practice a Japanese concept called Watashikai, I always take the initiative to introduce myself to those I Idolise. I also challenge myself to step out of my comfort zone, speaking to professionals in their local language and practicing cheki-style photography to capture moments of great importance. Enabled companies by improving their technological posture during COVID: adopting a hybrid cloud architecture! I employ the yakkai method in teaching, guiding and mentoring my fellow budding eventers step by step using visual and audio aids. I am learning wotagei in my free time, a cultural dance that connects me with my roots. Built a platform to allow developers to deploy a high churn application environment. Created Observability, Deployability and Robustness for a cross-regional DLT application on AKS for Financial institutions and their trade partners. Current Interests: Promoting SDGs, colouring the world Ultra Orange
- Achieved and maintained a Silver ranking for the GPU Cloud ClusterMAX™ Rating System 1.0 and 2.0 from SemiAnalysis, contributed in the scope of observability - Researched on and maintained production level observability for networking, thermal and hardware for A100, H100, H200 using NVIDIA Data Center GPU Manager and other tools for over 4000 GPUs. - Researched on and maintained production level observability for thermal, hardware and leak detection for NVL72 GB300 Compute Tray, Powershelf and Nvlink Switches; Inband using NVIDIA Data Center GPU Manager and Out-of-Band using Redfish APIs for over 18000 GPUs. - Researched on and maintained maintained production level observability in networking for Spectrum-X SN5610, SN2201 using NMX-T and GNMI metrics. - Created an exporter for slurm to allow to track the jobs user submitted, as well as utilization of each job. - Maintained an exporter that exposes transceiver metrics via mlxlink and processes lspci information; taking into account GPU passthrough. - Designed a global resilient and scalable observability architecture, covering 7 data centers across 4 geographical locations
- SCTP Cloud Infrastructure Engineering course for Adult Learners - Guided over 300 non technical learners through a structured program with Linux, Kubernetes, AWS, Cloud Architecture Design, CICD content - Mentored and inspired job seekers for their first step in their career change - Conducted knowledge sharing sessions to enable learners to spearhead digital transformations in their workplace, utilizing AWS and platform engineering practices. - Designed guidelines for students to create portfolio projects, challenging them to gain insights into real world applications - Created a interactive Kubernetes lab environment to teach learners compute, networking, monitoring, logging and storage concepts in a structured format
- Manage installation and upgrade of Kubernetes distributions on on-premise servers, scaling to tens of nodes. - Increase bid win-rate for application by 30% by reducing network latency by 600x, employing eBPF; implementing modern performance optimisations such as XDP load balancer service forwarding acceleration, Direct Server Return and BBR congestion control in Cilium. - Secured Kubernetes clusters by implementing CNPs, including services of kafka, grpc, rest types.
- Coordinated across Product and Engineering teams to roll out 4 production releases of the Contour (Corda - DLT) application on Kubernetes, generating $2M in revenue. - Optimized Java 8 springboot application on cgroups 1 and 2, enabling developers by introducing profiling and dashboards, reducing average application requests and limits by 30% in production. - Designed and implemented monitoring system including Thanos, Prometheus and Grafana across multiple regions to have a single pane of glass for application and system observability. - Constantly optimized cloud cost by right-sizing Kubernetes worker nodes and other cloud resources' capacity, saving $20K yearly. - Set up CVE Patch and SAST management and fix processes, for both OSS Kubernetes infrastructure components and application code. This ensured that security issues were addressed within 5 working days. - Set up a high churn Kubernetes environment for Developer team to deploy multiple instances of a private blockchain network. This allowed developers to safely and reliably deploy software to make high impact changes frequently and predictably with minimal toil at scale. Time to Market was decreased from 6 to 3 months. - Commercially implemented OWASP 3.2 ruleset, reducing garbage requests to application by 90%. - Worked with team to achieve ISO 27001:2013.
- Implemented landing zones in Azure, AWS and GCP for banks, local firms and MNCs alike; shifting applications to the cloud in less than 3 months.