Oxygen addresses key data management challenges with a future-proof data strategy

In the past few years, Oxygen’s business growth has created new data patterns across multiple departments, including finance, risk and fraud, marketing, and more. However, such growth presented its own set of challenges:

Oxygen used centralized cloud-based storage (Amazon Redshift) to store data from various departments and visualized it for business intelligence (BI) with Amazon QuickSight.

Gaps in the process became evident as the company started to bring in data from many different sources, including over 17 outside systems like payment tools and marketing software:

• The data architecture design created when the company was in its infancy couldn’t handle the increasing amount of data.
• No procedures were defined to guarantee data quality, accuracy, and compliance, and there was no designated ownership for each dataset.

As a result, the company faced occasional hurdles due to insufficient data management, affecting:

• decision-making processes
• reporting accuracy, risk and compliance
• information security
• business agility and change management
• manual upkeep of systems

These challenges blocked Oxygen’s ability to efficiently manage data workflows and adapt to changing business needs. Recognizing the need for a better data infrastructure to support scalability, security, and strategic agility amid rapid growth, the company sought a technology partner to proactively tackle these issues.

After evaluating Oxygen’s current data setup, Yalantis experts recognized the need to introduce a flexible data structure and governance capable of adapting as needs evolved.

This called for a future-proof data strategy, including:

• scaling the existing data architecture to handle increased volumes
• implementing a robust data governance framework to oversee data activities

With these objectives, the Yalantis team outlined key stages for the ongoing collaboration:

STAGE #1: AUDITING DATA ARCHITECTURE TO FIND GAPS AND OPPORTUNITIES 

Yalantis worked with the Oxygen team to audit and document the current environment, which included:

• mapping the current architecture — documenting key systems, data models, pipelines, dependencies, and integrations to identify areas of duplication, redundancy, and fragmentation
• reviewing infrastructure — assessing performance, capacity, security, costs, and gaps in analytics, reporting, and visualization capabilities to identify improvement opportunities
• quantifying issues — detailing technical debt, risks, and problems within the existing architecture to prioritize and guide remediation efforts
• fostering collaboration between DevOps, engineers, and testers to seamlessly transit to a scalable data architecture with established data governance processes
• discovering ownership of data processes to establish accountability for compliance and security

STAGE #2: IMPROVING REPORTING BY OPTIMIZING DATA WORKFLOW MONITORING AND MANAGEMENT

To create a more streamlined and governed data environment as Oxygen scales, Yalantis:

• mapped business processes with relevant data to ensure an automated supply of accurate data
• categorized datasets by business domain (operations, sales, marketing, etc.) rather than by technical task to let business users access data in the context they need for analysis
• automated and standardized data changes and updates by using GitLab to set up CI/CD pipelines
• optimized data storage needs and costs by eliminating duplicate or redundant data across multiple tables
• improved data visualization by adjusting BI dashboards in QuickSight to ensure accurate data views and reporting

STAGE #3: ESTABLISHING DATA GOVERNANCE FOR ONGOING DATA MANAGEMENT

Finally, since Oxygen deals with sensitive customer data, ensuring compliance, maintaining data quality, and building customer trust were crucial. That’s why Yalantis helped them implement a data governance framework focused on:

• defining data governance roles and responsibilities, such as data stewards and data owners, to provide accountability for data quality and security and mitigate the risk of data errors and inconsistencies
• implementing role-based access controls and encryption to securely transmit and store sensitive data like account numbers
• establishing KPIs to continuously monitor data quality and compliance risks and validate the accuracy, completeness, and consistency of reporting data
• conducting training to increase awareness of data governance protocols among employees