Feb 2020 – Aug 2021 · 1 yr 6 mos · London, United Kingdom

• Building Kore Operate, an in-house self-service multi-cloud (AWS, Azure, GCP) application platform using Kubernetes, Crossplane, Prometheus, external-dns, certmanager, and many more.
• I'm translating years of cloud infrastructure management and platform-as-a-service experience into Kubernetes Operators written in Golang.

Nov 2018 – Apr 2019 · 5 mos · London, Greater London, United Kingdom

• I managed the PaaS (Platform as a Service) Cloud Foundry service used by several government services. I provided training for junior team members and created education materials about Cloud Foundry.
• Used technologies/products: AWS, Terraform, Docker, Bosh, Cloud Foundry, Concourse, Open Service Broker API, Prometheus
• My role included:
• Monitoring (Prometheus), debugging and maintaining Cloud Foundry in multiple environments (using BOSH)
• Regularly updating Cloud Foundry using cf-deployment
• Writing and maintaining service brokers based on the Open Service Broker API using Go
• Extending the CI/CD workflow in Concourse
• Maintaining Docker images used by Concourse
• Maintain the AWS infrastructure using Terraform
• Regular pair programming
• Strict development processes including mandatory reviews and approvals

Oct 2017 – Jun 2018 · 8 mos · London, United Kingdom

• As part of the GOV.UK PaaS (Platform as a Service) team I managed the PaaS Cloud Foundry service used by several government services.
• Used technologies/products: AWS, Terraform, Docker, Bosh, Cloud Foundry, Concourse, Open Service Broker API
• My role included:
• Monitoring, debugging and maintaining Cloud Foundry in multiple environments
• Regularly updating Cloud Foundry using cf-deployment
• Writing and maintaining service brokers based on the Open Service Broker API using Go
• Extending the CI/CD workflow in Concourse
• Maintaining Docker images used be Concourse
• Maintain the AWS infrastructure using Terraform
• Regular pair programming
• Strict development processes including mandatory reviews and approvals

Jul 2016 – Mar 2017 · 8 mos · London, United Kingdom

• I managed the AWS infrastructure, CI/CD workflows for GOV.UK Notify and helped the Digital Marketplace team.
• Used technologies: AWS, Terraform, Ansible, Packer, Docker, Jenkins, Cloud Foundry.
• I managed the AWS infrastructure, CI/CD workflows for GOV.UK Notify and helped the Digital Marketplace team.
• Fully manage multiple environments with Terraform
• Move all server management code to Ansible
• Security hardening on Debian servers and on AWS (for pen-testing)
• Immutable server images with Ansible and Packer
• Migrate the Python-based Notify services to GaaP (Government-as-a-Platform), which was based on Cloud Foundry and set up a CD pipeline
• Introduce secret management for Digital Marketplace

Mar 2015 – Jun 2016 · 1 yr 3 mos · London, United Kingdom

• As part of the infrastructure team we built the whole infrastructure from the ground up, including a full CD pipeline, testing and local development environments. Our biggest challenge was running our services in China behind the firewall.
• managing AWS resources in three regions: Ireland, Singapore and China (mainly EC2, RDS, ElastiCache, CloudFront, Route53, Lambda, S3, SNS, SQS, VPC)
• managing four environments (development, testing, staging, and production) in AWS including multiple AWS accounts
• managing infrastructure and EC2 servers (>100) with Ansible (Debian)
• building Docker images for every application
• manage the complete CD pipeline for the student.com website
• creating fully automated testing environments for developers in AWS, where on single separated EC2 instances they could run together almost all services (using our existing Ansible playbooks)
• migrating from Ansible Tower to Rundeck
• setting up Jackrabbit replication for three AWS regions
• recreating the Vagrant development box including all applications with Docker Compose

Jul 2013 – Nov 2014 · 1 yr 4 mos · Budapest

• I worked closely with the Ustream Media Server which is a Java based distributed system to handle millions of viewer connections. Everyday development includes build configuration (Gradle), testing (JUnit, Mockito), deployment through multiple staging clusters (Vagrant, Chef), extensive measurement (Graphite, StatsD) and 0-24 monitoring (Nagios, Dashing).
• I wrote an open-source general log processing tool (Yolo) in Java to report events to graphite or statsd.
• I managed PHP components (like viewer routing and our internal API) which closely communicate with the media server.

Oct 2012 – Jul 2013 · 9 mos · Budapest

• I managed the backend team's daily tasks for 10 months - the team had five developers and a QA tester.
• As a team leader my responsibilities were to plan and organize our agile development cycles closely working with the Engineering Manager and the Core Systems Manager. We held daily stand-ups and regular team meetings to discuss everyday technical challenges. I took part in regular team evaluations.

Feb 2010 – Sep 2012 · 2 yrs 7 mos · Budapest

• I have built a build and deployment system from scratch (PHP, gearman). I have worked on the internal API, Ustream site localization, stream statistics processing and several other projects.

Nov 2007 – Feb 2010 · 2 yrs 3 mos · Budapest

• I was one of the developers of blog.hu which is a PHP and MySQL based, distributed blog publishing system serving more than 1 million PI a day.
• We wrote a PHP MVC framework from scratch (it contained a persistence framework with automatic caching and chain invalidation).