Ange Luvari

- migration of all the legacy app on AWS - setup serverless architecture with ecs, fargate, redis serverless and aurora serverless clusters - load-testing to understand and solve all the existing bottlenecks to handle a lot of users on a short period of time - setup terraform and back-end CICDs - setup of our own gitlab runners - finops studies and setup to reduce all useless costs

First year: - Development of the CICD used by all tech teams: - Multi-language (php/symfony, VueJs, Typescript, Rust, Python, C++) - Multi-type of projects (api, jobs, front-ends, daemons, libraries) - SysAdmin with server management on OVH Second and Third year: - Transition to squad teams: Project management with Jira (Sprints) - Docs management with Confluence - Transition from OVH to AWS - Setup of AWS services: - EC2: setup of shared/dedicated ec2 instances based on criticality with multiple environments - S3: setup of buckets for different uses (storage for front-end, data acquisition, static assets, etc.) - IAM: account security and access management - Cloudwatch: log/metrics management - RDS + redshift: database management - Cloudfront: front-end deployment - Route53: DNS management - SES: Setup email server - SNS: slack notifications - SQS: Setup fifo queues - SSM ParameterStore/SecretManager: credentials management - ECR/ECS: docker image management - api setup Fourth year: - Transition to tribes: Jira refactor to move to Kanban, Transition from Confluence to Notion - Transition to infra as code: - Study and creation of a battle plan to migrate to infra as code with comparative study of Terraform VS Cloudformation - Refactor of all legacy to move all infrastructure pieces to Terraform - Creation of Terraform stacks and modules - Management of certain automations with Ansible - Theory on AWS multi-account architecture - Use of Datadog Fifth year: - Transition from tribe to a Platform team - Transition from EC2 to Kubernetes: - Application and event auto-scaling with Keda - Node auto-scaling with Karpenter - Blue/Green deployment - Metrics/alerting with Prometheus and Grafana - Argocd - Helm - Setup OpScimBridge between OnePassword and Okta for credentials management - Refactor of cicd to gitflow or trunkbase depending on context - Setup OpsGenie for redirecting alerts for on-call shifts