Clip’s Journey from Data-Scrambled to Data-Fueled

About the Company

Mexico’s Clip is the leading financial integral ecosystem of its country. It was born with the mission to help businesses have a platform for digital payments. It provides mobile payment services, allowing businesses and consumers to make transactions by turning their mobile devices into card terminals. Clip serves customers worldwide. They promote the financial inclusion of people and businesses through innovative & technologically trusted solutions, making it easy, accessible, and transparent for all.

Challenge

Clip was aided in understanding what a data architecture modernization would entail while evaluating what the best steps would be across different options like a Data Warehouse, a Data Lake, or Lakehouse.

At the same time, Clip wanted to strengthen its systems for data replication from multiple sources and rethink how they did data modelling.

Additionally, Clip's Finance team was looking for help to automate their reports, consolidate their data sources into a Single Source of Truth (SSOT) and be able to reduce time spent in manual input, processing and validation of data. Mutt Data answered this call. Clip agreed on a Discovery Phase, which would allow us to assess their current pain points and business needs, and assess their current data architecture and analytical capabilities. The objective was to find a suitable and fine-tuned solution for Clip’s specific needs.

Considering the challenge, Mutt’s involvement in the project yielded deep findings and facts about Clip’s Data Structure, Data Sources, processes and needs.

In the eight weeks this Discovery Phase took, Mutt evaluated Clip’s pain points, technical debt, and scope while taking a close look at the technologies, architectures, and processes that were already in place.

During the Discovery Phase and while listening to the client, Mutt proposed different alternatives in Data Architecture, vendors for that data architecture, a clear blueprint, and a roadmap for implementation.

Mutt drafted and proposed the best-tailored solution for Clip aligned to their necessities and pain points, given the size of their business, the volume of data, and their data handling.

Mutt wanted the solution to meet specific KPIs while helping integrate multiple sources of data into an SSOT, data quality and improving performance.

Solution

During our swift Discovery Phase, we decided to engineer a solution that would involve the migration of Data Warehouse to Data Lakes. Mutt also found limitations in the platform providing the Data Warehouse so another platform needed to be brought to the table. Mutt also discovered the need to improve access control and handling of Personally Identifiable Information. Furthermore, we found an urgency to improve data modeling and orchestration.

After thoroughly reviewing the current architecture and aligning the proposal to the decision to use a specific data platform, the necessity of building a Data Lake using a multi-hop architecture as a foundation became evident.

In this context, it was decided to standardize each layer to enable data source integrations, which can apply data quality and transform it as they advance layer by layer.

This architecture will enable the following features:

• Multiple data source integrations, including ones that are not being replicated today in the warehouse, into a SSOT.
• Data quality control guarantees consistency.
• Data governance and sensitive information access.
• More robust and flexible process orchestration.
• Easier, simple, and reusable pipelines.
• Revaluation of data modeling.
• Improved performance.
• Reduced costs.

The modular and open design of the Data Lakehouse allows it to easily evolve the architecture according to what Clip may need in the future.

Impact

Once the Discovery Phase was completed, we were able to find, pinpoint and study solutions for Clip’s pain points. Although there are no clear-cut measurable metrics, the finding of each of Clip’s needs highlights what should be measured on the work done during implementation. Something very important to highlight is that the solutions proposed for the problems found are strictly tailored to Clip’s needs, there are not canned or one-size-fits-all solutions, but rather solutions tailored to their specific needs.