On Infosys Health Insights Platform. Q&A with Eric Paternoster

Eric Paternoster, President and CEO, Infosys Public Services.

Q1. At a high level, what is the Infosys Health Insights Platform?

Infosys Health Insights Platform (IHIP) is an automated data-science platform that enables organizations to turn their data into actions. 

IHIP aggregates, manages and analyzes data, irrespective of formats and types and delivers AI-driven next best actions that lead to more robust decision and policy making. 

With IHIP, organizations can become data-driven and use this single platform to support various program areas, like managing outbreaks, addressing the drug [and opioid] abuse crises, studying the impact of social determinants of health, and developing a clear view of the overall health (clinical, behavioral and mental) of populations.

IHIP will also spur collaboration, enabling multiple groups of public health workers including analysts, data scientists, researchers, epidemiologists, caseworkers, and policymakers to work together to deliver proactive and specific interventions to the constituents they serve.

Q2. So this is a solution Infosys has built in collaboration with AWS, Couchbase, and Knowi. What are the specific roles each partner plays?

Yes. All of us have combined our decades of experience in the public sector, data management, cloud and advanced data science to build the Infosys Health Insights Platform.

Couchbase provides the most advanced NoSQL engine and a modern data management architecture to the platform. From rapid aggregation of different types of data to multi-dimensional scaling and intuitive search capabilities, Couchbase enables IHIP to use the growing volume of unstructured and semi-structured data for lightning fast performance, real time insights, and persistence.

Knowi provides a unified analytics component that seamlessly connects with Couchbase and enables IHIP to analyze and translate data from multiple sources into meaningful visualizations. Knowi’s search-based analytics capabilities enables users of the platform to simply ask questions of their data and receive actionable data and visualizations right away.

AWS, in addition to the secure cloud infrastructure, provides data orchestration, analytics and security services, which reduce the infrastructure procurement and development time for agencies by as much as half and ensure that the platform is scalable. 

Q3. What were some of the drivers behind the development of the Health Insights Platform? What are the challenges healthcare organizations face that IHIP is solving for?

Government health agencies have always had access to a lot of data like the data sets from the Health Data Initiative, population health data, Social Determinants of Health, caseworker notes and other sources. But a significant portion of this data is unstructured. As such, agencies have only been able to analyze a fraction of the available data sets. This affects the quality of insights that they can generate. Take COVID-19 for example. The pandemic has established how poor our data infrastructure is.

The other challenge has been the amount of time spent to generate these insights. Despite access to different types of tools and solutions, analysts still spend almost 80% of their time preparing data. Given the extensive data prep required, it is difficult for agencies to generate the right insights at the right time. 

Aside from the issues with data and the inadequacy of the current tools, the shortage of analytics and data-science skillets in the market is a nearly insurmountable obstacle faced by many agencies. 

IHIP addresses all these challenges. It allows agencies to ingest, manage and analyze data of any type and format. IHIP automates the entire data analytics lifecycle – from ingestion and harmonization of data to developing machine learning/AI models and generating actionable recommendations. This makes the entire analytics process extremely efficient and effective. 

This single platform can be used for multiple program areas. Almost anyone with basic knowledge of analytics and statistics can use the platform to build advanced AI models, minimizing the dependence on experts and democratizing data science skills.

Q4. It seems that this model can be applied to other use cases. Are there plans to take this to other program areas?

Yes. The platform has been architected to support any program area across any agency/department. Currently, we have built use-cases around COVID-19 tracking and population health management and are in process of developing additional use cases for child welfare, Medicaid and social program eligibility and enrollment analytics. Over the next few months, we plan to add use-cases to support transportation agencies and commercial healthcare organizations as well. 

Q5. With so many data management vendors in the market today, what was the impetus behind partnering with Couchbase?  

There are multiple factors that led to the partnership between Infosys Public Services and Couchbase.

Couchbase is the most advanced NoSQL datastore in the market for persisting operational data. It offers a memory first architecture, support for horizontal scalability with low latency responses and an always-on architecture to support mission critical business systems.

Couchbase’s layer consolidated platform allows users to ask questions or query operational data in various ways. It could be through the SQL like query language or by asking full text search-based or analytical questions over large datasets.   Regardless, Couchbase provides the flexibility and extensibility to ask all types of questions without moving the data outside the operational database.

Couchbase’s strong support for microservice based workloads deployed on platforms like Kubernetes, OpenShift via its Autonomous Operator offering was a critical rationale for our partnership and additionally to support the consumption API’s that the Health Insights Platform was offers.

Q6. Can you describe some of the challenges you see with legacy RDBMS in health systems and how you’ve optimized data modeling and master data management by moving to Couchbase?

Traditional RDBMS systems struggle to support the requirement of scale and low latency response times that modern digital systems demand. Couchbase addresses these concerns with its memory first architecture, horizontal scale and always-on database design.

Health systems deal with data that is in different formats and susceptible to constant change. RDBMS’s impose a lot of rigidity on the applications due to the inherent schema restrictions. 

Couchbase with its flexible JSON schema allows organizations to model their database schema that is driven by access patterns. This gives health applications the agility and flexibility needed to handle multi-format and multi-type data. 

JSON schema allows organizations the flexibility to model their schema in a de-normalized, normalized or a custom manner to support application access pattern requirements. Combined with this, its N1QL query language engine, exposes a SQL like query language that helps existing developers with SQL language skills to quickly start developing apps on Couchbase.

Q7. With Couchbase, can you discuss how you’ve expanded the use and value of Couchbase for use in Machine Learning and AI use cases?  How does Couchbase’s analytic service play a role?

Couchbase allows an application, or an AL/ML based model, to query operational data in many ways. These can be N1QL queries, full text search-based questions or analytical questions.

Couchbase offers a HTAP (Hybrid transactional, analytical processing) system. It allows users to ask questions in different ways, in near real-time, without having to move operational data to other specialized system. This also allows organizations to save on the cost associated with moving data in and out of a cloud service. This supports the AI/ML and analytical requirements of the Infosys Health Insights Platform perfectly.

And let’s not forget, the Couchbase platform does all this without impacting the core operational performance because of the workload isolation architecture principle.


News: Infosys Announces Automated Data Science Platform to Support Public Health Agencies, Oct 07, 2020, 12:00 ET

Infosys will be presenting a demo of the new Health Insights Platform on October 15 at 1:00 p.m. Pacific Time during Couchbase Connect.ONLINE

Sponsored by Couchbase.

You may also like...