How to Overcome AI Roadblocks When Patient Data Is Scattered
“Stop talking about AI and focus on where your data is and what problem you are solving,” advises an industry expert.
Many health plans find themselves in the early phases of adopting artificial intelligence (AI). The challenge isn't the technology itself—it’s the data. AI requires consistent, reliable data, yet healthcare data often sits scattered across numerous systems. Kevin Deutsch, SVP of Health Plans at Softheon, emphasizes that to create real value with AI, health plans must first unify their data by prioritizing governance and regulation.
Are Health Plans Leading AI Adoption in Member Experience?
Healthcare overall is slow in adopting AI. Health plans aren’t necessarily ahead of providers. There’s a lot of talk about AI without a clear strategy focused on specific problems. The real hurdle is aligning AI efforts with business goals and addressing data architecture.
Key Challenges from the Member Perspective
Members want more personalized experiences. With shifts like the Individual Coverage Health Reimbursement Arrangement (ICHRA) and the blend of employer and individual exchange markets, a one-size-fits-all approach no longer works. AI holds promise to help health plans deliver this personalization effectively.
Current Impact of AI in Healthcare
AI is making a difference primarily in customer service. More chatbots and guided digital experiences are replacing traditional enrollment processes. Softheon, for example, uses internal language models to boost efficiency, delivering consistent, real-time answers to member inquiries. This frees up staff to handle more complex issues, improving overall service.
Engagement with Different AI Types
Softheon works with algorithmic, generative, and agentic AI. Their platform uses thousands of microservices, a form of agentic AI, optimized for specific tasks. Generative AI speeds up internal tasks like testing, generating test cases automatically to improve code quality. Access to extensive data enables these efforts. Additionally, health plans should consider applying AI to internal documentation and standard operating procedures to improve response times and accuracy.
Data Sharing with Providers
Softheon hosts data on a private cloud and offers APIs so carriers can access and store data on their own platforms. However, a challenge remains if other vendors don’t have the same level of access, preventing centralized data storage and effective AI application. Data is also shared through flat files and standard EDI transactions, but real-time access is increasingly demanded.
Real-Time Data and Provider Interactions
Healthcare data exchange is moving toward real-time streaming rather than batch EDI files. Real-time data is essential because any delay means outdated information. For AI to be effective, it needs current, accurate data on demographics, payments, and eligibility.
AI’s Role in Empowering Staff and Members
The first impact of AI is on internal teams, making their work more efficient and accurate, which in turn enhances member experience. The focus should be on augmenting staff capabilities, not replacing them. As one CEO puts it, AI replaces humans who don’t use AI. This mindset helps overcome hesitation around job automation.
Managing Scattered Data
Start by inventorying all integration points and identifying where data lives. Critical data should be stored in a structured, centralized location accessible in real-time. Not all data needs to be centralized—sensitive data handled by vendors might stay separate—but membership, service area, and operational data must be available instantly for AI to deliver value.
Member Acquisition and Billing with AI
- Member Acquisition: Use AI to segment populations precisely and deliver personalized experiences. Knowing your audience enables targeted communication beyond traditional methods.
- Billing: Ensure billing accuracy, timeliness, and ease of payment. AI can help provide consistent billing experiences, reducing errors and improving member satisfaction.
Future Outlook for AI in Healthcare
The distinction between human and AI interactions will blur. Soon, consumers won’t necessarily know—or care—if they’re engaging with a human or an AI agent, as long as their needs are met effectively. Personalization will increase, and human involvement may become minimal or invisible in many interactions.
Before jumping into AI, organizations must first identify where their data resides and clarify the problems they want to solve. Focus on data and problem-solving first; AI is a tool to help achieve these goals.
For healthcare professionals interested in practical AI education, resources like Complete AI Training's latest courses offer targeted learning to build relevant skills.
Your membership also unlocks:
