Why AI matters at this scale

Select Health is a Utah-based, non-profit health insurance company serving members across the Intermountain West. Founded in 1983 and employing between 1,001-5,000 people, it operates as a community-focused health plan, managing risk, administering benefits, processing claims, and building provider networks. Its core mission is to provide affordable, accessible healthcare, which inherently involves balancing member care quality with financial sustainability in a complex regulatory environment.

For a mid-sized regional insurer like Select Health, AI is not a futuristic luxury but a strategic imperative for competitive survival and mission fulfillment. At this scale, companies have accumulated decades of structured claims and clinical data—a rich asset for machine learning—but often lack the vast R&D budgets of national giants. AI offers a force multiplier, enabling Select Health to punch above its weight by automating manual processes, deriving deeper insights from its data, and personalizing member engagement. This directly addresses critical sector pressures: relentlessly rising medical costs, member and provider demand for seamless digital experiences, and administrative inefficiency that can consume 15-20% of premium dollars.

Concrete AI Opportunities with ROI Framing

1. Automating Prior Authorization: This is a prime target. Using Natural Language Processing (NLP) to review clinical notes and automatically approve routine, guideline-based requests can cut processing time from days to minutes. ROI comes from reduced administrative labor, decreased provider frustration (leading to better network relations), and faster access to care for members, potentially improving outcomes.

2. Predictive Population Health Management: Machine learning models can analyze historical claims, pharmacy data, and social determinants of health to identify members at highest risk for diabetes complications, heart failure admissions, or avoidable ER visits. Proactive nurse-led outreach can then prevent these costly events. The ROI is direct medical cost savings, improved quality metrics, and enhanced member health—core to a non-profit's mission.

3. Intelligent Claims Integrity: AI, particularly computer vision for document extraction and NLP for clinical context, can automate the review of complex claims. It can flag coding errors, potential fraud, or necessary medical record reviews with high accuracy. This reduces manual audit costs, accelerates clean claim payment, and ensures appropriate reimbursement, protecting the plan's financial health.

Deployment Risks for the 1001-5000 Size Band

Implementation at this scale carries distinct risks. First, integration debt: Legacy core administration systems (e.g., claims, enrollment) are often monolithic and difficult to integrate with modern AI APIs, requiring significant middleware or phased replacement. Second, talent scarcity: Attracting and retaining scarce data scientists and ML engineers is challenging against tech giants and well-funded startups, necessitating partnerships or upskilling programs. Third, change management: With thousands of employees, rolling out AI that changes workflows (e.g., for claims processors or care managers) requires extensive training and communication to ensure adoption and mitigate workforce anxiety. Finally, regulatory vigilance: As a health plan, any AI model making care-related decisions (like prior auth) must be rigorously validated for fairness, bias, and explainability to satisfy state regulators and maintain trust.