Healthcare sector

Maximizing savings by streamlining the procurement of medication across the hospital network

CUSTOM BACKEND BIG DATA

Creating a new version of an existing stock control solution, to process millions of medical records from external supplying systems within seconds?
Challenge Accepted!

Maximizing savings by streamlining the procurement of medication across the hospital network

Our client wanted to upgrade their existing stock control system for better efficiency and extended capabilities for importing, storing, and synchronizing data between the platform and external systems.

Through the streamlined management of product stocks, while following business best practices and government regulations, the client was able to make substantial savings. Group purchases of medication allowed them to channel more funds towards providing better quality care for patients.

Process & Story

Problem

The client needed a system that would maximize savings by optimizing product purchases through better control of stocks, the application of business rules, and full compliance with government regulations. It involved importing, storing, and processing millions of supply records from multiple external systems.

Most importantly, records in the system needed to be updated in real-time. This can be increasingly challenging with the growth of big data. Because the existing solution was entirely based on SQL Server Integration Services, it had some limitations. New and more complex business rules, along with constant data growth, forced our client to reach for a new approach.

Solution

The existing solution was upgraded to one based on microservices and leveraging the cloud. The goal for the new version was to create a solution that is testable, easy to maintain, and performs well under huge loads. Because the processing flow consists of particular ‘steps’, we separated existing logic and created a microservice for each one.

Data from files is divided into chunks that are sent and processed between services. However, they can’t be processed completely independently as records are connected to logical sets. To fulfill that requirement, we developed an internal balancer that ensures messages are processed within the logical sets and with respect to given resources such as a database.

The new system is much more efficient and easier to develop and maintain. The onboarding process for new clients is also now much easier and faster.

Case Study Schema Maximizing savings by streamlining the procurement of medication across the hospital network

Tools

Azure

.Net"

MSSQL

Azure Functions

Service Bus

Have a similar project in mind?

LET'S TALK!

Challenges

Ingesting data from multiple sources
It was prerequisite that the system quickly integrates new data sources. We unified the way incoming data is processed by standardizing the format and excluding differences in configuration modules.

Rapid processing speed is essential
The client needed to process data in seconds or minutes instead of hours and days. Compared to the old approach, we achieved almost real-time data processing speeds.

Missing data and errors
To more efficiently handle the huge amounts of ingested data and avoid potential processing issues (missing data, errors in code), we needed to implement a pre-processing stage. This allowed us to separate actual data processing from ingestion. First, the raw data is sent to a database, and only then it’s queued for pre-processing. In case of any issues in the pre-processing stage, we can quickly resume processing – starting from the records which generated issues, which saves a lot of time in the process.

Monitoring the system and correlating events
Proper monitoring is a must for a system that processes such amounts of data. We introduced complex logging and monitoring tools so the system can be tracked in terms of general usage, numbers of exceptions, or even rows processed by a particular service. The alerting system handles unusual situations and sends proper notifications.

Our other projects: