Alan Vitale, Vice President, Technology Solutions, Sagility
LinkedIn: Alan Vitale
Don Searing, Vice President of Product & Engineering
LinkedIn: Donald Searing, PhD
LinkedIn: Sagility
In the digital age of healthcare, data is abundant, but structure is not. Payer organizations are inundated with clinical notes, scanned forms, faxes, and handwritten appeals. These “unstructured” data types – which lack uniformity – slow operations, obscure insights, and increase administrative burden. Each type impacts payers in a variety of ways; none of them good. They create provider and member abrasion, leading to suboptimal experiences across the board with internal and external stakeholders.
But solving these challenges takes more than just great technology. It requires the right combination of intelligent automation and human expertise, working together to streamline intake, unlock insights, and ensure context isn’t lost along the way. With technologies like AI NLP (natural language processing) and ICP (intelligent content processing), paired with well-trained teams to manage exceptions and interpret data nuances, unstructured content goes from a challenge to a strategic advantage.
Modern payers with a strategic vision are reading between the lines – quite literally – to unlock smarter, faster operations.
The Problem Hidden in Plain Sight
Every day, payers process thousands of documents that don’t follow standard formats – think medical records, scanned forms, handwritten provider notes, or faxed appeal letters. Unlike structured data, like claim codes, this content doesn’t fit neatly into rows and columns. Rather, it requires human review, slows workflows, and often hides key information – such as urgency and reason – that’s critical for timely decisions.
In fact, more than 80% of healthcare data is unstructured – and most of it is underutilized, according to the Journal of Healthcare Engineering.
When exploring unstructured data technologies, payers should evaluate:
- Document diversity: What formats and sources are most common?
- Integration: How easily can these technologies interface with core systems like Facets or HealthEdge?
- Accuracy: Are AI models trained on healthcare-specific content?
- Scalability: Can the ICP/NLP solution scale with your data volume, handle increasing complexity, and adapt as your organization grows and evolves?
Traditional Systems Fall Short
While digital transformation has touched many parts of the payer ecosystem, legacy systems still struggle with the foundational challenge of handling unstructured data efficiently. And, in general, healthcare is notoriously slow in adopting new technologies, like AI.
Traditional technology systems are frequently outdated, run on old software and hardware, and, basically, aren’t up to the task of running through data effectively.
Legacy systems often fail to efficiently manage:
- Medical records
- Faxed claims and appeals
- Handwritten notes or PDFs
- Free-text fields in forms
- Emails or voice transcripts
These documents require manual intake, routing, and interpretation, adding friction to processes like grievance and appeals, clinical reviews, or case management.
Enter ICP and NLP
To truly transform operations, modern payers must embrace ICP and NLP technologies to automate the extraction and support interpretation of unstructured content. With an experienced team managing the ICP and NLP work, this tech-enabled process reduces the amount of provider and member abrasion caused by older, slower legacy systems while improving business operations.
ICP uses a combination of OCR (Optical Character Recognition), machine learning, and workflow automation to:
- Read documents (scanned or handwritten)
- Identify document types (referral vs. appeal)
- Extract key fields (member ID, reason for denial)
- Route to the correct department or system
For health plans ready to move to next-level document processing with even higher performance, NLP makes it possible by interpreting context and meaning.
Some particularly useful capabilities of NLPs include the ability to understand sentiment or urgency in text, recognize clinical terminology and intent, and extracts structured data from narrative notes.
The People + Tech Equation
Even the most advanced technologies can’t solve complex healthcare problems alone. Successful payer operations depend not just on automation, but also on the expertise and scalability of human support teams. A well-orchestrated operation integrates front-office and back-office professionals alongside AI, ensuring exceptions are handled, escalations are resolved, and member-provider experiences remain seamless and positive.
Having a dedicated workforce trained to manage non-automated exceptions, validate extracted data, and provide contextual understanding ensures that AI systems operate not in isolation, but in partnership with people. This human-tech synergy is not just valuable, it’s essential. Technology drives efficiency, but people ensure empathy, context, and trust remain part of the equation.
Real-World Example: Automating Grievance Intake
Let’s say a health plan receives a handwritten member grievance about a delayed authorization. Previously, using the typical legacy system, the complaint required scanning, reading, and manual entry into a tracking system.
Using an ICP+NLP pipeline completely revitalizes the old process:
- The document is scanned and digitized.
- NLP identifies the complaint type and urgency.
- Key fields (member ID, issue description, date) are extracted automatically.
- Grievance is routed to the appropriate person (with a summarized and annotated document) in real-time.
The result? Faster resolution, better compliance tracking, reduced human error, less stakeholder abrasion, and better internal and external experiences.
Efficiency, Benefits & Beyond
Payers that capitalize on the technologies available today and anticipate those arriving in the future will have the ability to outperform and outpace their competitors. Further, implementing intelligent solutions creates an output that is more than the sum of its parts.
The technology described supports improvements that can lead to:
- Reduced manual document handling of 40%–60%
- Improved turnaround times for grievances and appeals
- Better audit readiness and traceability
- Enhanced member and provider satisfaction
Most importantly, it unlocks insight-rich data previously trapped in static files – fueling analytics, trend detection, and smarter planning.
In the past – and for those payers still mired in legacy systems – unstructured content was a roadblock. Today, it’s an opportunity. With technologies like OCR and NLP, supported by skilled operational teams, payers can finally make sense of the noise, turning scattered data into structured action. When AI and people are aligned, efficiency improves, friction decreases, and outcomes accelerate.
In a healthcare landscape driven by speed, precision, and personalization, that human + tech partnership makes all the difference. Not every exception can be coded, and not every decision should be automated.