This site uses cookies to offer you a better browsing experience. By continuing and navigating through the site, you accept the use of cookies.

For full information, see our Privacy Policy.

An American Mass Media Company
Transforming Data Management
a computer screen with data

Goals:

● Serve as a central self-service portal and repository for specific analytical data

● Allow for retrieval and analysis of data across the organization 

● Cater to different stakeholders’ data needs and address varying levels of technical proficiency

● Enable both quick business insights and deeper data science initiatives

Challenges:

● Current data exists in silos across the organization

● Retrieval is not self-service 

● Background and definitions require human interaction

Architecture Overview

The architecture behind the Analytics Platform is made of custom components built on top of Cloud-based services and products. These components are responsible for managing a variety of data manipulation steps, such as extraction, transformation, and load, among others. They are built in such a way that they can be easily reused throughout the platform and connected to each other in order to create more complex pipelines.

You can think of these components as independent services that migrate data from one place to another or that transform from one format into another format. The output of a step can be the input of another step and so on. Usually, all the pipelines are built so the data ends in a Cloud-based Data Warehouse solution, which enables businesses to query and work on top of that data—the main business goal of the Analytics Platform.

 

Workstream Pillars

1. Data Management 

The goal was to establish "data process" routines and platforms separated from "data access" methods and applications to increase readiness for analytic platform capabilities—Date Science & Machine Learning. 

2. Business Intelligence and Analytics

We enabled self-service for data visualization using presentation layer tools like Tableau. We build a layer where minor items can be turned around quickly to enable any given business requirement. 

3. Automation and Continuous Integration 

We established CI-CD driven data extraction, transformation, and load routines in conjunction with deploying automated testing, error tracking, and alerts for optimized performance. 

4. Cloud Infrastructure 

We recommended establishing a Cloud infrastructure the was both economical and competitive. It provides a scalable infrastructure and high-performance ETL and Business Intelligence layers. 

flow chart