Problem Description

A technology company leader in Property and Casualty Insurance Claims reached out to Allata to help them optimize their Build and Deploy process for claims and billing software, which was currently being performed manually. 

This technology company develops claims and billing software intended for property and casualty insurance. They offer custom solutions, as well as integrations with leading industry vendors, to enable clients to capitalize their business and solve critical challenges.

Before partnering with Allata, this company’s manual Deployment process involved several key tasks. Once the Artifact was ready, they had to set up a VPN connection, connect the Remote Desktop to customers’ Windows Servers (On Premise or Azure Virtual Machines), copy files and finally, deploy the Internet Information Services website. Also SQL scripts had to be executed manually. This entire release application process (build and deploy) took several hours, was error-prone and relied on one person who was responsible for this task.

Solution

Given that this technology company was already using Azure Repos and Azure Boards as part of the Azure DevOps product, Allata proposed and implemented a CI/CD solution for the release application process using Azure Pipelines (among other features of Azure DevOps). The idea was that this solution would better integrate with other features of Azure DevOps that were already being used by the client.

• Azure Repos were configured as triggers for the Building process.
• Azure Pipelines were used to Build the application software depending on branch rules.
• Deployment Groups were created and used to release applications through the Azure Agent (PowerShell script) installation on Customer Servers.
• Azure Release Pipelines were used to deploy the application and execute SQL scripts on Customer environments (Dev/UAT/Prod).
• Notifications about Build and Release were sent to Slack and Teams channels to track the process results (fail and success).
• By using Service Connections, Azure DevOps was connected to Gitlab. Through this connection, Azure Pipelines were able to successfully build code from Gitlab repos.

Benefits

By using Azure Pipelines as a CI /CD tool, the customer was able to take advantage of key benefits, such as:

• Faster release of applications (Build and Deploy).
• Scheduled releases on specific days/times.
• Double-check security: releases configured to wait for approval in order to prevent unwanted deployments.
• Eliminated manual Builds and Deployments, therefore eliminating error-prone activities due to human errors.
• No third party dependencies such as VPN and RDP: the Azure Agent running on customer Servers was responsible for directly connecting to Azure Devops and performing all the required tasks (Artifact download and deployment).
• Notifications on Slack and Teams about success or fail on Build and Release Pipelines.
• Automated tracking of every operations deployment with Azure Dashboards.
• Eliminated single-point-of-failure dependencies: no longer relying on a single person dedicated to perform manual Builds and Deploys. 
• Freed up time for the person dedicated to performing manual build and deployment: now they could focus on management responsibilities, team tasks and processes.
• The Azure Repos were migrated to Gitlab after a couple of months. By using Service Connections, the Azure Pipelines were able to continue Building from Gitlab repos/branches.