Migrating an Inventory Management System Used by 1,600 Enterprises to a Cloud

Summary

Brief results of the collaboration:

• Zero-downtime migration to the cloud enabled the company to innovate its legacy on-premises system, taking the burden of installation, maintenance, and support from 1,600 enterprises.
• The shift to a modular architecture allowed for selling the platform as a whole product or by individual components.
• The allocation of computing resources though Microsoft Azure's elastic pools stabilized performance during load spikes and helped to save up to $12,000 per year on infrastructure provisioning.
• CI/CD pipelines facilitated feature deployment to just 6 minutes.
• A custom authentication system, role-based access control, SSL encryption, identity/app/device management ensured proper security.

The customer

Based in Norway, the company is one of the leading providers of inventory management software/hardware. Founded in 1994, the customer serves both public sector and private enterprises.

The need

The company offered an inventory system to manage assets (e.g., office equipment), purchase- sale agreements, invoices, etc. This was an on-premises product, so users had to install/ maintain it by themselves. It also put a strain on delivering technical support. Multiple workflows, such as creating a catalogue of items, remained manual. So, the customer decided to migrate to the cloud. As the existing solution was built on WPF, Microsoft Azure was chosen to retain the codebase as much as possible.

Lacking relevant in-house talent, the provider relied on .NET development expertise at Altoros to build a multitenant system focusing on automation, as well as long-term maintenance and support.

The challenges

When working on this project, the team at Altoros faced the following challenges:

• Due to the multitenant nature of the product, it was important to prevent load spikes from impacting performance.
• Since the system stored sensitive data, security was a priority.
• It was crucial to ensure no data loss and zero downtime during migration from an on-premises to a cloud-based system.

The solution

Stage 1. After analyzing the requirements, .NET developers at Altoros devised a roadmap and prioritized tasks. Following the Agile methodology, the team worked in frequent iterations.

Stage 2. For the ease of maintenance, the team designed a modular architecture. This also enabled the customer to sell the system either as a whole or by individual components (lending, documentation, licensing, etc.).

Stage 3. Using built-in functionality of Azure SQL Database, engineers at Altoros created elastic pools to allocate computing resources based on the load. This helped to ensure high availability and stabilize performance during spikes. On top, it reduced expenses on infrastructure provisioning.

Stage 4. For security, the developers built a custom authentication system, implemented role-based access control and SSL encryption, as well as integrated Microsoft Intune for identity, app, and device management.

Stage 5. To facilitate migration, the team at Altoros automated database creation for tenants in a cloud system. Then, the engineers moved data from the legacy solution, keeping structure and all the dependencies up-to-date. The developers also wrote validation scripts to prevent data loss, duplicates, etc.

Stage 6. The engineers created templates to automate data import/export for reporting and invoicing. The team at Altoros also integrated Active Directory and configured advanced filters/data mapping for inventory tracking.

Stage 7. CI/CD pipelines with Azure DevOps facilitated feature deployment to just 6 minutes.

Stage 8. After migrating to the cloud, Altoros provided ongoing maintenance and support.

The outcome

Partnering with Altoros, the customer migrated its inventory management system—utilized by 1,600 enterprises—to the cloud. Thanks to a modular architecture, the system can now be sold both as a whole or by individual components. The allocation of computing resources via elastic pools not only ensured performance during load spikes, but also helped to save up to $12,000 per year on infrastructure provisioning. Satisfied with the collaboration, the company plans to further expand the functionality of the platform and introduce more automation.