Connor Atchison’s disability claim, the one that ended his military career, took 18 months to resolve. That timeline was not unusual —many claims extend far longer — but what stayed with Atchison was how much of that delay came down to process. Information arrived in fragments. Critical details were buried in hundreds of pages. Decisions depended on how well someone could piece that together under time pressure.
He then spent the next seven years in healthcare administration, working on the other side of that same process. The pattern held.
Information was still fragmented, key details still buried, and decisions still relied on someone piecing everything together under pressure.
“On one side, someone’s waiting for answers. On the other, someone’s trying to make the right call without the full picture.”
Seeing both sides brought the issue into focus. Claims don’t stall because of a lack of urgency. They stall because the underlying information is difficult to interpret, organize, and trust at scale.
That realization became Wisedocs.
Claims operations have improved at the margins without changing how decisions are made. When paper became digital, workflows became more structured. Outsourcing was able to absorb volume. At the same time, complexity increased.
Medical records now arrive in multiple formats, from multiple providers, often with inconsistent timelines. Claims teams are expected to process that information quickly and accurately, even as experienced adjusters leave the workforce and are not easily replaced.
“You are dealing with more volume, more diversity in documents, and less knowledge transfer,” Atchison said. “And the process still has to be accurate every time.”
Accuracy is not a secondary concern. A missed detail can alter the outcome of a claim, extend its duration and create reputational risk that is difficult to reverse. Claims decisions are among the few moments where insurers are judged directly by the people they serve. Once credibility is questioned, it is difficult to rebuild.
The work itself adds pressure. Reviewing large files requires sustained attention, often across thousands of pages. “No one wants to spend an entire day on a single file if it can be done faster with the same or better quality,” he said.
The industry’s response has been to layer in tools. Claims document management systems, automation of medical chronologies and AI-based summarization have reduced manual effort and review times by up to 80%.
These tools help as they standardize inputs and reduce repetitive work. But they do not resolve the central issue, which is how decisions are supported.
“There is always a new tool,” Atchison said. “But a tool on its own does not fix the process.”
The distinction that matters is between processing information and making intelligent decisions. Summaries can be generated quickly; however conclusions still need to be defended.
In claims, that standard is non-negotiable. Decisions must be explainable and auditable. Systems that cannot show how an outcome was reached introduce a different kind of risk.
“If it is a black box, it is not defensible,” he said. Research backs this up: trust in AI-generated outputs increases fourfold when expert human oversight is added to the process — a finding that underscores why accountability cannot be automated away.
A more durable model is taking shape, one that combines structured workflows with targeted automation.
Certain tasks can be handled efficiently by machines, such as document intake, chronology building, identifying inconsistencies and flagging anomalies. Platforms like Wisedocs are built around this layer, using AI trained on large volumes of claims data to extract and organize key information while maintaining human oversight.
The role of the human expert does not disappear. It shifts.
“You want to automate the repetitive work and elevate the expert,” Atchison said.
That shift matters because claims decisions require context and judgment. They also require a clear chain of reasoning that can be reviewed later. AI can surface the relevant details. It cannot replace accountability.
Within claims speed, cost and accuracy are linked, improving one often pressures the others. What is changing is the ability to compress time without weakening the outcome.
Tasks that were outsourced with two-week turnaround times can be handled internally in near real time. Wisedocs' platform, for instance, delivers first touch on medical records 60 to 80% faster than manual review — with consistency that outsourcing could rarely guarantee. The gain is not only speed, it is consistency.
“Companies that can put that together quickly and do it the right way are going to pull ahead,” Atchison said. “The ones that cannot are going to struggle.”
As these systems become more embedded, the focus is shifting from capability to control.
Claims decisions must be traceable and outputs must be auditable. Models must operate within clear guardrails. A faster process that cannot explain itself introduces a different kind of risk.
“Defensibility and governance are going to be critical,” Atchison said.
That requirement will shape the next phase of adoption. It will also establish a clearer standard for what constitutes a usable system in a regulated environment. The industry has spent years trying to move faster. The challenge now is to move faster while maintaining clarity over how decisions are made.
For Atchison, that question is not abstract. It comes back to the experience that led him to build Wisedocs in the first place. Claims are not abstract workflows, they are sequences of decisions that affect real people, often at difficult moments.
When those decisions are wrong, or even perceived to be inconsistent, the cost is not limited to time or expense. It is credibility. And once credibility is compromised, it is difficult to restore.
This article was produced in partnership with Wisedocs
