Qovery has driven a 40% increase in delivery speed, a 25% reduction in costs, and a 35% boost in developer productivity at Julaya!

#1. Introduction

#Hi Mickaël, can you introduce yourself, and your main responsibilities at Julaya, please?

I'm the "I know a guy" who is always Swiss tools on all IT things 😅.

More seriously, my role is a multiple one (*ops) cause I have long and multiple experiences:

• Platform Engineer
• DevOps/DevSecOps
• Information Security Officer
• SRE
• Cloud Architect
• Solutions Architect
• Developer of course (always in my heart ^^)

My responsibilities are mainly the following:

• Infrastructure Cloud management (actually AWS & GCP), availability and performance
• Infrastructure Cloud Costs
• Continuous Integration and Delivery
• Automation (technical and transverse)
• Monitoring/Testing/Alerting management and performance
• Incident preparation and response
• Perform Infrastructure with architectured solutions for Infra & Services
• Developer Experience performance enhancement and make Product team happy
• Network, System, and Security manager in all offices (endpoints, mobiles, networks, VoIP, ...) with the help of my colleague recruited exclusively for that part
• Compliance for PCI DSS and SOC2 (preparation)
• Responsible for a budget (Infra/System, Networks/Security)
• Security: Data protection, Proactive security & response, Posture Management, culture, training…

But I must admit I'm very bad at making good coffee 😅

#2. Overview

#Can you briefly describe Julaya's infrastructure/deployment setup before implementing Qovery?

We have 3 AWS accounts for our services (product):

1. Live (prod)
2. Staging (pre-prod + Q&A)
3. Development (Q&A)

Each account is managed by Terraform (only by DevOps). Each service on each account has a CodePipeline which runs one or many CodeBuild and potentially deploys artifacts

All Q&A was made on development and then on staging before applying to the production environment. Code is versioned and backed by Github with GitHub Actions for continuous regression testing and future version approve

#3. Challenges

#Can you share the specific challenges or pain points you were facing in your development and deployment processes?

1. Firstly, the main problem was the QA experience.
   - Difficulty in having constant and no broken data in the development phase
   - Difficulty in having scoped data  in the development phase
   - Difficulty in having a scoped development branch with only the feature/bug to test due to constant merging in the development branch
2. Secondly, another problem was to perform better developer experience
   - Difficulty with the exponential microservices communications to test new features/bugs without breaking the development branch
   - Difficulty with the exponential use of AWS/3rd party services to test new features/bugs without breaking development branch
3. Thirdly, all the automation needed to create for a small DevOps team
   - Time to automate entire workflows and processes
   - Time to create/manage complete/complex system

The first goal was to improve our product and make Julaya employees more confident and happy.

#4. Qovery Implementation

#How has Qovery contributed to streamlining and automating your deployment process?

Here are Qovery's main contributions to our workflow:

• Scoped environment by feature/bugs/improvements/refacto
• Flexibility to create POC
• Leverage developers to do by themself
• Flexibility and enhancement of our QA
• Staging is no longer necessary (cost killing + speedest deployment workflow)
• K9 managed (for a small DevOps team is helping a lot)
• Easing by managed automation: Preview environment automation, Docker, and Helm facilities

#Can you report any improvements running Qovery in Production OR on your development workflow?

Yes, the impact has been surprisingly positive.

Regarding our development workflows:

• More speed in delivery time (+35%) by leveraging the Preview Environment feature
  - Better Q&A time
  - Better continuous integration
• Reduce versioning complexity by leveraging the Preview Environment feature
• More reliability:
  - Less bug deployed in production (-30%)
  - More reactivity on
• Reduce DevOps time (complete managed system with a lot of automation already done)
• Cost killing
  - Preview env ran only when needed
  - Complete deletion of staging environment (on next step so a potential gain of 5000$/year)
• Deployment time has been increased not due especially to Qovery, but with the type of jobs we running (AWS Lambda by CloudFormation) that take a long time (actually we working on this part to decrease deployment times)

#5. Results and Outcomes

#Any measurable outcomes, such as time savings, cost reduction, or increased developer productivity?

• Time Savings: ~40%
• Cost Reduction: ~25%
• Developer Productivity: ~35%

#5. Fina World

#If you can tell me one sentence regarding your overall experience with Qovery?

Qovery allowed us to almost effortlessly redefine new best practices for a better Developer Experience, save DevOps man time, and a better continuous integration and testing workflow, thus allowing us to increase our quality but also our velocity. A real success and it’s only the beginning!