BI Tableau platform for Data Management and Insights

Business Domain:
Healthcare
Service:
Data Engineering, Business Intelligence
Technologies:
Oracle, Java, Tableau

Business Overview

The client is a multinational SaaS provider focused on business intelligence solutions for the healthcare industry. The core product is a BI solution for building embedded analytics and implements extensive reporting and visualization functionality for healthcare organizations’ executives to track and find areas to improve financial, clinical, and quality of care indicators.

The client contracted NIX to boost the existing SaaS Tableau-based platform capabilities, enable more accurate data analysis and gain actionable business insights for healthcare organizations and their employees.

The NIX team scope included the following:

Expand end-users capabilities to customize interface and fine-tune dashboards based on existing business needs
Support multi-tenancy for dashboards and ensure data separation for each organization
Ensure secure user and client data synchronization with the Tableau platform

Solution

The existing configuration of Tableau didn’t provide enough customization options for the end-users’ business needs. NIX’s top-tier data engineers designed a custom visualization solution to widen Tableau’s capabilities and provided the following enhancements:

Tableau workbook extension

We designed a custom configuration component to embed an HTML/js-based user interface into the Tableau dashboard. It enables more flexible customization and generation of user-facing dashboards with cross-filtering and in-depth reviews of the organization’s metrics.
Multi-tenancy via Tableau

We used existing Tableau functionality to support multi-tenancy in the product. Each client was represented as a site in Tableau Server, which guaranteed the data and content separation deployed to each site.
User and client data synchronization

We developed a separate authorization service that ensures private and secure access for each tenant by storing the data in its own PostgreSQL database. To sync the data to Tableau, we developed an individual component that inputs the required information into the system.

Data management

Since the data in the product's database updates daily, NIX developed a supplementary Java-based component to schedule Tableau extract refresh tasks on the server every night to keep workbooks up-to-date. Input data from open sources or data sent by clients directly were loaded into a database daily.

The data is structured and has three metric types:

Clinical data allows comparing benchmarking organization performance against expected performance rates.
Financial data helps measure progress, establish targets and baseline current performance through validated benchmarking methodologies, and drive performance improvement initiatives.
Patient satisfaction data provides valuable information about patients experiences with hospitals, clinics, doctors, and nursing staff, such as overall quality, outcome of treatment, and more. This helps enhance patient satisfaction and workflow for the entire organization.

Each client’s database contains about one hundred tables with more than one billion rows each. Since the amount of data calculation and filtration heavily loaded the database, we orchestrated data separation in a cascade-like way to ensure the data filtering and processing fits the structure needed for Tableau visualizations. Each step allowed us to remove irrelevant data and shape it into the appropriate forms.

Outcome

The enhanced visualization capabilities and feature-rich interface of the SaaS platform help the client deliver a more personalized user experience and advanced business intelligence opportunities, which resulted in becoming one of the top-line healthcare providers on the market.

The delivered enhancements empowered healthcare organizations with data-driven performance measurements, leading to increased patient satisfaction and operational efficiency.

Tech stack:
Tableau, Backbone JS, Java, Spring Boot, Hibernate, Liquibase, Oracle, SQL and PL/SQL
Team: 11 experts
Tech lead, Business Analyst, 3 Java Developers, 3 QA Engineers, 3 Data Engineers