Client background
The client helps improve and facilitate clinical trials and studies by providing software solutions that help determine research data quality, accuracy, and integrity. The company has over 10 years of experience and a global clientele.
Published client profile: Industry: Healthcare and Life Sciences. Partnership period: December 2020 - present.
Business challenge
The client aimed to modernize their Big Data monitoring platform that analyzes clinical and operational data during trials and detects atypical data patterns (tempering, systemic errors, fraud, etc.). They wanted to re-architecture the solution to make it more scalable and flexible and enhance its functionality with new integrations and additional features. This would help improve the service quality of the client’s solution, help them attract new customers, and grow revenue.
Delivery approach
The delivery team assisted the client in designing a new, highly scalable infrastructure for the monitoring solution and all its additional features. The delivery team assessed the client’s existing architecture, which included comprehensive analyses of the monitoring solution, the administration platform, and the risk-based oversight application. During this assessment, the delivery team engineers:
Helped design the new architecture;
Analyzed data storage systems;
Identified challenges in the existing data architecture;
Determined business needs and requirements, such as migration to the cloud, data storage and management standardization, pre-production and production testing automation, and others.
Finally, the engineers helped the client conduct code refactoring and utilize a data-centric architecture approach while developing new microservices-based solutions.
Implementation
The delivery team helped the client re-architecture the monitoring platform, making it more scalable and allowing it to process larger volumes of data quickly and efficiently. The delivery team implemented a microservices-based Data Storage service and set up a data pipeline that ingests, processes, and transports data in the following stages:
Conversion of csv files to the Parquet format, which resulted in faster data transfer and reduced storage costs.
Transformation of data written in Python and SAS into a Parquet file, which is easier to analyze.
Creation of dataset snapshots which store data in various formats for several applications.
Storage of datasets on Azure blob storage instead of MySQL, which resulted in high scalability of the solution in terms of the amount of data it can store and the rate at which data can be accessed.
Data from the Data Storage service is rendered in the new medical study and review platform that the delivery team engineers helped develop. This platform contains interactive dashboards and charts that display insights on each clinical study. It allows data analysts to filter trial details by various attributes, leave comments, and assign reviewers to conduct deeper data analysis.
Furthermore, the delivery team helped automate the process of creating data cubes, which represent and analyze data across different characteristics, such as patient ID, region, country, etc. As a result, this accelerated the data analysis setup from approximately 2 hours to just under 3 minutes per customer.
The delivery team also helped automate action item generation for cases when data anomalies are detected. Instead of manually inspecting data and creating action items for deviations, customers can simply set up and configure specific thresholds. If data crosses the set threshold, the system automatically generates notifications to the customer’s CRM and ERP systems. This accelerated data anomaly detection from several days to just minutes.
Next, the delivery team assisted the client with manual and AQA testing of third-party integrations and system UI, as well as the setting up of new tools. Our AQA engineers helped unify automation tests across the monitoring platform and optimize end-to-end automation test cases. This reduced the test automation lifecycle from 24 to 3 hours and achieved a 90% test automation pass rate.
Moreover, the delivery team helped design and implement a performance testing solution that can launch multiple tests with different scenarios on numerous environments at the same time. This included implementing a single test reporting tool for all QA teams, which provides analytics for all unit tests and aggregates and compares results on a daily, weekly, or quarterly basis.
Together with the client, the delivery team implemented automated regression testing for multiple solution environments. The delivery team set up testing for system migration to a separate environment, which allowed us to verify changes in the database and architecture components without disturbing other environments, manual testing, or the development process.
Finally, the delivery team helped the client build a new separate administration platform, which was originally part of the monitoring platform. The delivery team migrated all its features from the monitoring platform, such as user and study management. The delivery team also established effective synchronization between the two systems.
Value delivered
The delivery team helped the client re-architecture the monitoring platform, which included providing a comprehensive architecture assessment, implementing and integrating new features and tools, and setting up automation tests. The delivery team also helped develop a separate medical study and review platform, and an administration platform. As a result of the engagement, the client gained several substantial benefits:
Increased customer base and engaged a Fortune Global 500 company by enhancing their services with a new monitoring platform that offers a wide range of useful features;
Grew their revenue by 115% in 2023 alone as a result of the increase in customer base;
Improved platform performance by up to 14.5 times by utilizing cloud-native features for storing data and applying a data-centric approach that allows the platform to be scaled easily.
Expertise delivered: Software QA & Testing, Big Data, DevOps consulting Technologies: Python, Angular, MS Azure, Flask, Pyramid, FastAPI, Celery, MySQL, MongoDB, SQLAlchemy, Cypress, Kubernetes 115% increase in revenue in 2023 Up to 14.5x improvement in platform performance Several days to a few minutes reduced data anomaly detection times 90% test automation pass rate
