The Challenge
Inter.ai team continuously develops and enhances its product with each iteration starting with a viable market-ready prototype.
They did not have a production system or an artifactory storage, the POC was manually deployed. Sphere was tasked with building a management server that would provide endpoint management as Inter.ai solution transitions towards the SAAS model.
Inter.ai also lacked an official build process and the new builds were manually deployed. Release deployments were stressful, because the new builds/ features were never tested together and the team encountered many bugs out of the blue, some due to wrong assumptions and others due to the environment.. The new builds were also being deployed as the latest version of the product and a unique management server was used for every user/client which resulted in higher cloud hosting and storage costs.
How It Was Solved
Sphere performed analysis of the present state of the Inter.ai infrastructure and their builds process, which was followed by analysis of the overall Continuous Integration/Delivery-Deployment strategy. Putting CI/CD in the cloud is a cost-effective solution that offers reduced maintenance cost, time management, and space (to house servers), resulting in greater business agility. Here are some of the line items where Sphere has already produced results for Inter.ai or are currently working on:
- Improved CI\CD pipelines, which increased flexibility, speed, and reduced bug count
- Optimized current AWS infrastructure, which reduced monthly costs for AWS services
- Implemented Dynamic Slaves for some of the projects in Jenkins, which increase the flexibility, speed and reduce costs for AWS EC2 instances
- Modified some testing workflows, which reduced AWS EC2 instances costs by eliminating unnecessary stages
- Helped migrate to the PostgreSQL database from Tiny DB, which increased speed, stability and scalability of the Management service
- Implemented monitoring service based on Prometheus + Grafana, that can be used later for microservices monitoring
- Implemented VPN in the company which increased the level of security
- Implemented Auto Source Code Signing mechanism (from CI pipeline), which will increase the speed and convenience of releasing new versions
- Implemented Active Directory (with Group Policies) and Okta, Full Disk encryption, and other security controls for the upcoming ISO 27001 Certification
- When there was a POC for CNN during Election 2020 in America, we helped with automation of their AWS infrastructure creation and management (whole infrastructure was based on AWS services)
The Results
As Inter.ai transitioned to a SAAS model, Sphere Partners built a management server, providing Inter.ai with a cost-effective CI/CD solution, effectively increasing flexibility, speed and bug count. By further eliminating unnecessary stages and modifying workflows, we reduced instances costs. We continue to work with this client, supporting them on their transition to a SAAS model.
