Aegis

Inspiration

Aegis AI was inspired by the realization that healthcare revenue is bleeding out because the current billing system is designed to be reactive. Our team learned that first-pass denial rates have climbed significantly since 2016, with current rates sitting at 10% to 20% , and a staggering 50% to 65% of those denied claims are never resubmitted.

What it does

Aegis AI is a desktop application that predicts insurance claim decisions and instructs providers on how to fix errors before submission. By simulating the payer's decision engine, the app provides a confidence score for approval, denial, or underpayment, allowing billing teams to address issues proactively.

How we built it

We developed a three-step processing pipeline: first, an ingestion engine reads CMS-1500 forms, PDFs, and medical notes. Next, we built an AI model that simulates the payer's decision engine to predict outcomes. Finally, we implemented an optimization layer that provides ranked fix recommendations to reduce denial risk before the claim is sent. The prototype of the model is built in the Figma. The data for the model can be implemented from previous denials and acceptances. That data is then used to build a confidence interval which can simulate the output from the insurance company.

Challenges we ran into

The primary challenge during the ideation phase was finding a way to meaningfully address the high frequency of errors in medical billing, as studies show 49% to 80% of bills contain at least one error or coding issue. We struggled to conceptualize a solution that could accurately simulate complex payer logic to justify the effort of a pre-submission rework.

Accomplishments that we're proud of

We are proud of conceptualizing a tool that transforms claim denials from a reactive burden into a proactive advantage. Our system successfully automates the identification of critical data points like CPT codes, diagnoses, and prior authorizations, which helps providers watch their denial risk drop in real-time.

What we learned

During Ideathon 2026, we learned that administrative friction is a major source of pure revenue loss in healthcare. We discovered that by providing proactive billing tools, we could help hospitals achieve faster payments and significantly reduce the high financial costs associated with manual rework.

What's next for Aegis

Our next step is to expand the conceptual AI's modeling capabilities to cover a wider range of specific private and government payers. We also plan to design deeper analytics to help hospital administrators track long-term revenue recovery and further refine the ranked recommendation engine for even higher accuracy.