About Me
Ramirez Ruiz
Sr. DevOps Engineer with a Bachelor of Science in Applied Mathematics, and a minor in Physics from the University of Central Arkansas. Skilled in Python, R, Swift, Kubernetes, Terraform, and Istio service mesh.
Sr. DevOps Engineer with a Bachelor of Science in Applied Mathematics, and a minor in Physics from the University of Central Arkansas. Skilled in Python, R, Swift, Kubernetes, Terraform, and Istio service mesh.
Factory Data (FD) is responsible for delivering advanced analytics, strategic insights, infrastructure as code, and algorithms to support our factory. FD integrates Machine Learning and Data Science into our manufacturing facilities, providing a competitive edge.
Key Projects
(2023): I led the efforts to migrate from Terraform to Terragrunt to ensure all our Infrastructure as Code would scale with ease. We currently only have 1 factory, but with this setup we have the infrastructure to scale globally.
(2022): As the dedicated DevOps Engineer supporting the new Drive Unit Shop, known as Enduro, I ensured that all new stations were seamlessly connected to our networks, enabling data to flow into our Manufacturing Data Warehouse from day one. This included creating custom deployments in Kubernetes to support moving large data from Kistler Servo Presses and Vision stations. I also assisted with deploying LineConfig, an internal tool used to setup every station (used by MFG Engineers to configure Enduro stations).
(2022): I built Snowgate & SnowFTP, internal programs that stream data directly into Snowflake from various sources like NATS clusters (messaging system similar to Kafka), AWS Kinesis, and ML applications like Petra-AI into our data warehouse.
(2021): I served as the key developer for the FD-Filesync project, aimed at moving images and video files out of the factory floor into our data warehouse. FD-Filesync moves more than 336 terabytes a year. This program is used factory wide and is leveraged heavily in the following shops: Battery, Body, Drive Unit, End of Line, and General Assembly.
(2021): Responsible for building and maintaining the Manufacturing EKS clusters, which power internal applications used across the factory.
Skills/Software
Terraform, Terragrunt, AWS, Kubernetes, PostgreSQL, Redis, MySQL, ETL Pipelines, Fivetran, Gitlab CI/CD, Kubeflow, Git, Python, MLflow, Kubeflow, ZenML, Snowflake, and more.
During my tenure at Rakuten, I served as a DevOps Engineer, bridging the gap between the DevOps and Data Science teams. My role primarily involved managing the development, staging, and production environments for the Rakuten Catalog Platform team. My core responsibilities encompassed a wide range of tasks, including system configurations, application lifecycles, infrastructure deployments, and the continuous development of internal tools designed for security audits.
At that time, I had the responsibility of overseeing 60 Google Kubernetes Engine (GKE) clusters of varying sizes. I was directly accountable for clusters used for training machine learning models and clusters housing Elasticsearch databases which fed data to over 1 million jobs per day.
Beyond my routine DevOps work, I led several significant efforts during my time at Rakuten. I orchestrated a major network migration that impacted all existing workloads, facilitating the transition to a service mesh architecture by moving to an Istio Service Mesh. Additionally, I collaborated closely with several Google Cloud engineers to test and debug Alpha programs, with one of the most notable being the A100 NVIDIA GPU program. We were the first Google Cloud customers to use A100 in GKE. I also played a key role in guiding my Data Science team towards adopting Infrastructure as Code (IaC).
Lastly, I successfully facilitated the deployment of Kubeflow into production by implementing private clusters, utilizing Google Cloud Storage (GCS), and integrating CloudSQL. This architecture enabled future upgrades to be performed smoothly and securely, eliminating the risk of data loss.
During my tenure at Metia, I worked in a small Data Science team building unsupervised Natural Language Processing (NLP) Topic Models written in R and Python for Microsoft executives. During my tenure, I was able to optimize the models to run in several hours as opposed to several weeks. I helped productionalize a web crawler that assisted with data collection for our models, and maintained numerous databases.
Oct. 2018 - Jul. 2019While working at Euronet Software Solutions, I helped maintain several production environments for banks located in Australia, Sri Lanka, and Latin America. I helped with failing over to DR environments when necessary and assisted with production failures. My core responsibility was to ensure banking systems remained active and continued processing transactions.
Dec. 2017 - Oct. 2018When I interned at Entegrity, I lead an energy rate analysis consisting of interactive energy maps. These maps assisted the business with identifying potential new locations for future expansions. I also setup the building blocks for their auditing application named enForm, that was later developed by me.
Dec. 2016 - Dec. 2017
enForm an iOS application, designed to help engineers process equipment data faster and efficiently. enForm has many features that range from multithreading, saving data offline, continuously syncing equipment data, and passing workloads to worker nodes provisioned on Google Kubernetes Engine (GKE). GKE ensures that the services are always available, and additional containers can be deployed to process the data being requested. By taking advantage of autoscaling, costs are kept at a minimum whenever these services are not in use. Recently enForm and the Entegrity team was featured in Energy News turning solar savings into increased teacher pay in Arkansas.
Ramona's Radiant Rooms allow users to make payments, schedule new appointments, and retrieve their payment history. All payments are made using tokens, and all the transactions are encrypted. The user can save as many debit cards, and credit cards as they would like. The cards are not stored locally, instead they are stored in Stripe's backend, and all the information needed is fetched using tokens. This provides extra layers of security keeping all the card information off the devices. This application allows Ramona's Radiant Rooms to reach more clients and process payments securely.
Google Kubernetes Engine (GKE) & Elastic Kubernetes Service (EKS)
Network Architecture
Cloud Security
Logging - Prometheus, ELK, Stackdriver
Continuous Integration and Continuous Delivery (CI/CD)
Python
R
Swift
C++
Java
Kubeflow
Spinnaker
Jenkins
Internal Deployment Tools - Python
Terraform
Istio - Service Mesh
Natural Language Processing (NLP) - Topic Models
iOS Development - iPhone, iPad
Machine Learning
Databases - Bigquery, Elasticsearch, Firebase, Prometheus, SQL, Rabbit MQ, Redis
© 2023 Royer Ramirez Ruiz