In partnership with SAANS Health, we managed ambiguities and engineering complexities with continuous design, user testing, technology spikes, and agile execution to streamline the launch of this InsurTech marketplace.

Partnership Context

SAANS Health has a proprietary AI/ML platform, AiQ, where augmented intelligence is an accelerator to bring healthcare products to the market. An InsurTech company was using Care and Risk AiQ to build a new insurance marketplace. This insurance marketplace would empower small and medium businesses to negotiate cost-optimized insurance plans.

Three competencies were required to create and launch a next generation platform: design thinking, rapid prototyping and user testing, and end-to-end development.

The constituents involved in developing the solution:

• Insurance industry product and domain expertise from our InsurTech client
• Solution definition, data and ML expertise, and industry expertise from SAANS Health
• Zero-to-one acceleration and ambiguity-handling expertise from Zemoso Labs

SAANS Health’s CEO, Raju Kattumenu, is a successful serial entrepreneur and has partnered with us on his previous ventures to accelerate go-to-market. He recognized that we would be able to help productize his health-data solution for this InsurTech use case in a time-boxed, expeditious manner.

Partnership Priorities

The table stakes included standard prototyping, user testing, user onboarding, and MVP launch milestones. But, the primary goal where Zemoso’s expertise played the most important role was to bring all these expertise together, align, identify what really mattered to the end-user, and launch the MVP in 6 months.

This would require us to hone in on the following three priorities:

• Integrate SAANS Health AiQ within the InsurTech platform so that it could be leveraged seamlessly
• Solve the every-evolving complexities of carrier integrations to extract/retrieve claims data through SMBs
• Set up protocols to reinforce user authentication around sensitive data, and continuous HIPAA compliance

Apart from the above, we of course helped design and build the marketplace for both insurance providers and SMBs, automated communication of benefits, etc.

‍

Time-boxing everything

To get to the outcome of an MVP launch in 4-6 months, we time-boxed every milestone. Our stakeholders were active participants throughout the journey, helping us align, make decisions, and move forward. Depending on the week’s priorities: the agile pods adapted, changed configuration, and focused on the outcome.

• Conduct 5-day Google Ventures Design Sprint remotely for key user flows
• Co-created high-fidelity prototype in four weeks
• Created an incredibly detailed design system to ensure execution consistency in 6 weeks
• Conducted user testing, and implemented UI feedback in weekly sprints
• Rolled out new aspects of the features in bi-weekly sprints

Prioritizing the Essentials and the Complexities

System basics
We ran an architecture sprint to create a logical and technical architecture. Some heuristic decisions were made over tech-stack choices. The platform leverages the modular benefits of a microservices architecture. The front-end was built using React JS, and using atomic design principles. Backend was built using Spring Boot. QA automation, CI/CD, etc. were of course built into the delivery lifecycle.  

The core of the platform: Data‍

Key challenge: Member-mediated data access

Access to Health Data for SMBs is critical to their ability to negotiate and procure cost-advantaged health insurance. At the core of the platform is a data retrieval system used to fetch member data from insurance carrier portals. This entire process is member-mediated. Selenium-based robotic process automation and secure API interfaces were used to interface and retrieve data on behalf of the member. AiQ analyzes this extracted data and provides reports to SMBs to help them negotiate a better deal for their employees. Insurance vendors use these insights to provide informed proposals. We also used RabbitMQ messaging layer to queue the data extraction process, and in doing so, the platform could balance the load optimally. ‍

The crux of the solution - a marketplace of disaggregated insurance products

Key technical challenge : field matching and alignment

The workflow for the SMB marketplace required vendors to submit their proposals into the system. Due to the variability in the inputs from the various vendors, we incorporated a fuzzy search mechanism to create consistency in the representation of the various quotes to the end user, so that they can easily compare and select.

Security measures - encrypted at rest, and access-controlled

Access to the sensitive health data of an individual from within an insurance carrier’s database is member-controlled. Okta gates access to this health data (from any other insurance provider or employer) within the InsurTech platform with strict user authentication protocols.    

The secrets management process stores the member credentials in the database in an encrypted state. This is decrypted only when the Selenium bot needs to run the API call into the insurance provider’s database. All member information for access to API-based data sources is stored in GitHub Secrets, ensuring end-to-end security.

Mapping system for the database

A mapping system was built to give developers more control over the data retrieval and transformation process. This allowed for consistency of processing through the AiQ platform.

Conclusion

In the end, it was an irrefutable, all-encompassing win as we helped the InsurTech client onboard their first customers and pressure test what we’ve built together.