Tiger Analytics spearheads migrations to Azure and Snowflake

October 29, 2025

Organizations need to adopt more flexible, scalable, and cost-effective solutions to keep up with the evolving landscape of data analytics. Traditional on-premises systems, while robust and reliable, often face limitations in handling the increasing volume and velocity of data. To thrive, organizations must modernize their critical applications, replacing legacy systems with cloud-native solutions that offer better scalability, fault tolerance, and security.

Microsoft partner Tiger Analytics is at the forefront of this modernization process. The organization has all three Solutions Partner designations for Azure, as well as three specializations: AI Platform on Microsoft Azure (formerly AI and Machine Learning on Microsoft Azure), Build AI Apps on Microsoft Azure, and Analytics on Microsoft Azure. Tiger Analytics used their expertise to execute a digital transformation for a leading global intimates and sleepwear retailer’s analytics ecosystem, delivering a seamless migration from the on-premises model to Azure and Snowflake.

Over a period of 10 months, the intimates retailer was able to transition out of legacy on-premises systems with a 100% success rate on all workloads—and significantly reduced costs along the way. The framework Tiger Analytics used to execute this migration can be tweaked for any organization based on their priorities, processes, and current ecosystem.

Collaborating for a smooth migration

Executing the migration in two phases

Tiger Analytics executed the migration in two phases. First was the migration of noncritical workloads, which involved the refactoring and rebuilding of applications from on-premises/SAS to cloud-native technologies. Each module's SAS scripts were meticulously converted to Snowpark, ensuring that the logic and functionality were preserved. Tiger Analytics utilized bespoke accelerators to speed up code refactoring and conversion on the most used SAS procedures and functionalities. Then came building pipelines and automation, preparing for cutover, testing the converted scripts, and optimizing performance.

For the second phase, they migrated critical workloads. This followed a similar approach to the previous phase but focused on higher impact modules and SAS Stored Processes. This phase ensured that all analytical workloads were converted and validated before the final cutover. During this phase, the remaining SAS scripts were converted to Snowpark, and each converted module underwent continued testing and validation. SAS Stored Processes were redesigned to Streamlit apps that run on Azure App Services, hosted on the intimates retailer’s custom domain and backed by stringent security protocols. Then, modules that passed validation were transitioned to the production environment, ensuring that all workloads were running smoothly on Snowpark.

Once all modules were successfully migrated and replicated in the new environment, Tiger Analytics oversaw the sunsetting of the old platform and services, which was executed without affecting business operations.

This phased migration strategy ensured a systematic and controlled transition from on-premises to cloud, minimizing risks and disruptions while maximizing the benefits of the new analytics environment.

Inside the streamlined architecture

The migration from on-premises to cloud introduced a streamlined and efficient architecture that leveraged Azure services for scalable and secure setup while maximizing the use of Snowflake's powerful compute resources.

The new architecture had a few key components:

1. Azure VM instances
Setup and connectivity: VM instances on Azure were configured to connect to Snowflake over a public network. The IP addresses of these VMs were whitelisted on Snowflake to ensure secure and seamless connectivity.

Workload execution: These VMs were tasked with running analytics and reporting workloads. They effectively managed data processing tasks, leveraging Snowflake's compute power.

2. Snowflake compute virtual warehouses (VWs)
The architecture made extensive use of Snowflake’s VWs, which are scalable compute resources that can be dynamically adjusted based on workload requirements.

3. Application hosting
SAS Stored Processes to Streamlit apps: The SAS Stored Processes were converted to Streamlit applications, enhancing usability and integration with modern web technologies.

Hosting on Azure App Services: These Streamlit apps were containerized using Docker and hosted on Azure App Services, ensuring high availability and scalability. They were hosted on a custom domain to enable SSO and accessed through a private endpoint for added security.

A continuous commitment to success

Optimization and monitoring are not one-time activities, but ongoing processes that drive continuous improvement and innovation.

The global intimates and sleepwear retailer recognized the need to optimize and worked with Tiger Analytics to create a comprehensive dashboard. This dashboard, which was built using Streamlit and hosted on Azure App Services, provides close to real-time updates. It consolidates cost and compute consumption data, runtime performance benchmarks, failure events alerts, and an ongoing log of executed pipelines into one user-friendly dashboard.

The migration has set the stage for exciting opportunities to further enhance and optimize this setup. So far, significant benefits have been realized in the six months since migration, including a 34% reduction in compute costs, the establishment of a secure and scalable platform, and high adoption within the data science team to train, test, and deploy critical workloads on Azure and Snowflake. Building upon these successes and embracing cutting-edge technologies will enable a more sophisticated data analytics and machine learning ecosystem, which will in turn drive better and faster decision-making and an enhanced customer experience.