Stephen Johnson Email & Phone Number

Wales, Gb

London, London, Gb

Migrated platform from Heroku to AWS using Terraform, ECS, EKS (Kubernetes), Lambda, RDS, GitHub Actions, and CircleCI.Reduced platform spend by 30%, by right sizing AWS resources, moving services to pre-commits and moving to an enterprise discount plan. Responsible for designing, implementing and monitoring scalable and secure infrastructure on AWS.Maintained and built the Chinese infrastructure on AWS, aligning it with AWS Dublin infrastructure, working within Chinese restrictions and regulations.Migrated their platform from Ruby to Python, whilst moving from Circleci to GitHub actions. Responsible for Python tooling and build pipelineRan multiple security audits of the Platform

Cambridge, Gb

Managed a two-person team in a machine learning startup.Utilised EKS (Kubernetes), Terraform, AWS CDK, and ECS for maintaining the corpus and AI model deployments.Implemented large batch processing jobs using spot fleet instances to manage costs.

Boston, Massachusetts, Us

I managed a team that supports internal and external IT services. My managerial role consisted of setting targets and budgets quarterly, and the direct line management of four staff members, including career progress planning, one-to-ones and performance reviews. Most external services utilise the following technologies; AWS (E2C/ Lambda / S3 / ECS / EKS / KMS), Docker, Prometheus, Nexus, Jenkins, Kubernetes, Airflow, Vault Spark and Terraform, with most tooling developed in Python. While at Opensignal, I have rewritten the build and deployment pipelines for all of the Android, IOS, and web applications developed in-house. I have also migrated services to a serverless architecture or EKS to drive down costs and increase availability. This included multiple internal and external audits focusing on security and privacy. Opensignal focuses on aa data analytics platform, and uses a mixture of technologies to provide this Spark, Airflow, S3 and AWS Spot fleets. These technologies offer "big data" solutions to customers. I have been involved in a project to update how we best use Spot instances and run Kubernetes and Spark together to provide a more stable, cost-effective, and scalable solution.

Boston, Massachusetts, Us

San Francisco, Ca, Us

I managed a team of two to support the BIM360 organisation using a mixture of the following tools, AWS (E2C/ Lambda / S3 / Apigateway / ECS / KMS), Docker, Chef, Sensu, Artifactory, Jenkins, Kubernetes, Terraform with some Ruby, Python and GO development. Due to the massive cost savings and reliability increase, I also led a team that migrated from EC2 / Dockerized apps onto Lambda. I was also responsible for taking multiple teams through the SOC2 certification process during my time.

South Nutfield, Surrey, Gb

I managed a team of two to support the BIM360 organisation using a mixture of the following tools, AWS (E2C/ Lambda / S3 / Apigateway / ECS / KMS), Docker, Chef, Sensu, Artifactory, Jenkins, Kubernetes, Terraform with some Ruby, Python and GO development. Due to the massive cost savings and reliability increase, I also led a team that migrated from EC2 / Dockerized apps onto Lambda. I was also responsible for taking multiple teams through the SOC2 certification process during my time.

Minneapolis, Minnesota, Us

During my time at Puppetlabs I was responsible for evangelising Puppet by speaking at multiple conferences and community meetings worldwide. Solving problems with Puppet for companies of all sizes and industries this included automating their infrastructures using Puppet and Mcollective. As well as Consulting on DevOps best practices, agile methodology, streamlining processes, creating training materials and teaching classes. I also worked with a large bank to deploy Puppet / Mcollective into their environment to automate sever builds, service deployment, infrastructure hardening and monitoring.

At AHC my primary duties included designing systems and services for 3rd parties and deploying those services. All services were hosted at our collocation facility and had to be secured to industry standards such as PCI DSS etc. The primary technology stacks at AHC were RedHat Linux, Ruby, Passenger and Apache. All applications and servers were deployed using Puppet and monitored using Nagios.

My main duties whilst at Graze, was the setup and administration of the development, testing and production systems. This mainly utilized the use of configuration management utilities such as Puppet. As well as the monitoring / configuration of inter site communications over leased lines and VPNS (openvpn + ipsec). Graze’s entire web services run on the AWS platform the following software is employed at Graze, Apache, Nginx, Mysql, PHP, Memcache and the Zend Framework. Running on a base operating system of Ubuntu with all Graze software being packaged.One of the main projects whilst at Graze has been the automation of deploying both services and servers in order to lower mean time to market, as well as extensive work on Ganglia and Nagios infrastructure to ensure trending and monitoring are in place for capacity planning. Both AWS elastic loud balancer and Nginx are employed as load balancers, allowing Graze services to scale horizontally and improve reliance. All services use MYSQL (Percona) in Master / Master and Master / Slave configuration to increase redundancy and for scalability.During my time at Graze we also went through the process of becoming PCI/DSS certified, I was one of the main employees behind the program and luckily we could leverage Puppet and other tools to impose the necessary configuration changes to the infrastructure.

Espoo, Southern Finland, Fi

One of the main projects at Nokia, involved the design and building of a new datacenterto host development and production systems, that involved deployment of a Cisco / HP basednetwork, Sun Openstorage, HP MSA Storage, ESX, KVM and F5 load balancers. The primaryserver hardware used within the datecenter are HP DL460 G5/G6 blades and HP DL160 G6Servers. All system run on RHEL that is managed using puppet and installed usingkickstart. This is one of many datacenters that are currently used by our department tohost both production and development systems running on ESX and KVM. The main project that i worked on involved theautomated deployment of blade centres world wide using RHEL 5.5 and Red Hat Satellite server. This project will be replacingYahoo Maps service with one provided by Nokia and is using Akamai for global loadbalancing. In order to make deployments more manageable throughout the rest of the maps department,a system using puppet, rpm and RHEL was created that allowed for automated deployment andupgrading of both the operation system and applications. This also linked in with ourmonitoring services Nagios and Hyperic to automatically create the necessary monitoreditems as well as using Ganglia for trending. Another project I also worked on was thecreation of a Hadoop (20 node) cluster that was used for offline processing of data thatlater was pushed into the production systems. The main applications that I support run on the Tomcat and Jboss application servers andare load balanced using either Apache or F5 load balancers with Varnish proxies front-ends. These also use a mixture of back-ends this included MYSQL and CouchDB.

Au

Company moved offices to AmericaMy role at Boku Ltd was primarily that of design and implementation of new systems and processes within the company. As a start up company, the infrastructure needed a total redesign in order to provide fault tolerance and cross site redundancy. As well as this a new backup process, change management process, security policy and design policy had to be implemented. The applications where running on a mixture of Perl code and PHP code and using MySQL as a back- end storage system, running on RHEL 5 and Debian Linux. My other day to day responsibilities also involved the configuration and management of Bind 9 both for external and internal DNS resolution and configuration of exchange and active directory.As well as supporting the subversion / git repositories and TRAC ticket system used within development.

Bournemouth, Dorset, Gb

Headhunted to Boku by previous ManagerMy primary role within UFI ltd is to support and maintain the main learning environment that is used by over 2.2 million users per year. This involves not only insuring the service is up and running 24x7 but also taking preventative measures to circumvent any problems which may arise. The technologies that are employed within the current organisation range from Solaris10 running on a range of Sun platforms (sparc and x86) to RHEL 4/5 running on IBM/ Dell blade servers as well as a mixture of Windows 2000 and 2003 server. Within the organisation a mixture of web based technologies are used including .NET, JSP and PHP which all have to be supported. The services which I supported within UFI ltd are APACHE, Postfix, Mysql, Jboss, Tomcat and a variety other system. NFS, GFS and samba was also implemented within UFI to produce fault tolerant and distributed file systems to insure the maximum uptime for the services UFI offer. Every system is load balanced and fault tolerant using CISCO CSS’s and firewalled using a mixture of CISCO ASA's and Checkpoint firewalls. Due to the mixture of technologies we used, we have a wide variety of backend databases including Oracle, SQL Server (including version 8) and MySQL, all of which I support and manage on a daily basis

North Sydney, New South Wales, Au

The role involved integrating our systems with other online marketing companies around the world and this has allowed the search works to infiltrate the US market. With an estimated 20 million dollars worth of data utilizing the code that had been developed, this has allowed the search works to expand.My main duty whilst at TTW was the setup and administration of the development / testing systems. This involved setting up multiply development / testing system in order to ensure the product was ready for a production environment. This involved multiply testing system from FreeBSD, Linux and windows, as well as multiply technologies Apache, Mysql, PHP, Perl. Another major task is supporting a system which ran over multiple Apache and MYSQL clusters in order to handle the load on the system and also meet the SLA. With over four hundred clicks hits per second, any outages would have been unacceptable and would affect the companies standing in the online marketing sector.