One Health Link

Company Background

‍

One Health Link transforms healthcare journeys with AI-powered, personalized solutions that make healthcare easier, smarter, and more human-centered. Built on millions of real-world interactions, their platform drives sustained engagement and intelligently guides people through the care ecosystem to lower costs and improve access, quality, and outcomes. However, delivering this vision at scale required confronting systemic challenges across the entire healthcare ecosystem.

‍

Challenge

‍

Healthcare stakeholders face interconnected pain points. Health plans struggle with uncoordinated outreach that creates member confusion, diminishing trust and hindering access to benefits. Their impersonal communication lacks the self-service options consumers expect.

‍

Providers face post-COVID financial pressures from labor shortages and burnout. They lack cost-effective patient engagement methods, hindering their ability to ensure appointment attendance, treatment compliance, and preventative care adherence.

‍

Patients lack timely, personalized support to access and use their benefits—forced into an unwanted "choose your own adventure" through a fragmented healthcare system. They want help that neither payers nor providers can deliver cost-effectively.

‍

One Health Link needed an AI solution addressing these pain points—starting with the labor-intensive live chat workflow where call center agents manually typed responses to guide new Medicare members through a multi-step welcome onboarding process. Their existing RAG-based copilot was rudimentary, requiring agents to copy and paste AI suggestions between systems. They needed a solution that could auto-populate intelligent, contextual responses directly in the agent’s workflow while maintaining healthcare's accuracy and quality standards.

‍

Solution

‍

Stride built an AI-powered agent portal for One Health Link’s live chat call center agents, providing intelligent response suggestions that auto-populate directly in the agent’s workflow. Built with LangGraph for the AI agent backend, React TypeScript for the frontend, and Azure OpenAI for the model, the system operates on client-provided call scripts (“blueprints”), FAQs, and standard communication templates. The initial deployment focused on supporting agents who serve Medicare beneficiaries through a structured welcome onboarding process.

‍

When a member is escalated from the automated chatbot to live chat, the AI agent reads the full conversation history, identifies where the member is in the onboarding workflow, and auto-populates a suggested response directly in the call center agent’s text box. Each suggestion includes a confidence score with reasoning, source document references, and validation attributes. Human-in-the-loop oversight ensures agents review and approve every response before it reaches the member, maintaining accuracy for each situation.

‍

This design transforms call center agents into efficient quality gatekeepers. With AI-generated responses arriving in under ten seconds—optimized from initial response times of over five minutes—agents can review and approve suggestions far faster than manually composing answers. The agent maintains conversational context across multi-turn interactions using persistent threads, so it understands the full history of each member engagement. The AI handles information retrieval, blueprint navigation, and response generation while humans provide healthcare judgment, expertise, and final approval.

‍

The chatbot addresses systemic challenges: providing patients immediate guidance, helping payers coordinate outreach and improve engagement, and giving providers cost-effective communication tools that improve compliance and preventative care.

‍

Outcomes

• Massive Operational Efficiency – Reduced manual response times from 5 minutes to under 10 seconds, allowing care teams to scale without increasing headcount or burnout.

• Better Health Outcomes – Delivers near-immediate, personalized guidance to Medicare members, driving better appointment attendance and treatment compliance.
  
  
• Validated Scalability – Successfully delivered "Human-in-the-loop" AI model while establishing a production-ready Agent to expand on.

‍

‍