Breaking Through Healthcare's Most Challenging Frontier

On the latest episode of the AI in Chicago podcast, we had the privilege of speaking with Bobby Reddy, Jr., CEO and co-founder of Prenosis, about his decade-long journey to develop the first FDA-approved AI tool for sepsis diagnostics. The conversation revealed profound insights about innovation in acute care, the realities of building healthcare AI, and what it takes to solve truly complex problems.

The Acute Care Innovation Desert

One of the most striking revelations from our conversation was Bobby's explanation of why acute care has seen so little innovation over the past 30 years. The reasons form a perfect storm of systemic challenges:

Biological Complexity: Unlike other conditions, acute care patients arrive with multiple comorbidities and complications occurring simultaneously. As Bobby noted, "These patients can have acute heart failure and have kidney injury all at the same time. They could come in with COVID." This heterogeneity makes the biology extraordinarily difficult to understand.

Perverse Financial Incentives: The DRG (diagnosis-related groups) reimbursement system means hospitals get paid the same amount regardless of the technology they use. This removes financial incentives for innovation, forcing hospitals to pay out of pocket for new solutions.

Big Pharma Disinterest: Despite acute care representing 10 times the spending of all oncology combined, pharmaceutical companies view it as too difficult and too small a market to pursue.

The result? We've been spending enormous healthcare dollars on the sickest patients while investing the least in developing solutions for them.

The Power of Problem-First Thinking

Bobby's journey from academic researcher to CEO offers a masterclass in entrepreneurial evolution. Initially driven by the desire to start a company, he spent years building technology in search of an application. The turning point came when he gained unprecedented access to a hospital, spending two years talking to doctors, patients, nurses, and lab technicians.

"We really don't understand the biology of patients in the acute care setting," Bobby explained. "We've never run like a human genome project to understand patients in the acute care space."

This insight led to a fundamental shift from building a product to solving a problem—a distinction Bobby emphasized as crucial for any entrepreneur: "Don't focus on building a company. Don't focus on building a product. Make sure that at the core of why you get up every single day is to solve a real problem."

Building AI That Demands Better Data

Prenosis took an unconventional approach to AI development that challenges a common assumption in the field. Rather than accepting missing data and having algorithms impute values, Bobby's team built their sepsis immunoscore to refuse to operate without complete information.

"The dirty secret about AI is it doesn't do well with missing data," Bobby noted. "Most AI algorithms are just making up, they're imputing the missing data. The sepsis immunoscore, what we do is we actually force the hospital to order the missing data."

This approach resulted in a system that requires 22 specific parameters, including two biomarkers (procalcitonin and C-reactive protein) that hospitals don't typically order for these patients. By forcing comprehensive data collection, they achieved a 65% positive predictive value compared to the 13-15% typical of early warning systems.

Transparency as a Trust-Building Strategy

Perhaps most importantly, Prenosis built transparency into their AI system. Rather than creating a black box, they show clinicians all 22 parameters used in the calculation and which factors increase or decrease sepsis risk.

"It's really important for clinicians to forgive the tool," Bobby explained. "When it gets things wrong, it's very important for clinicians to be able to see why the immunoscore thought that patient was going to be septic... I forgive the tool because I see why it made the mistake."

This design philosophy acknowledges that AI systems aren't perfect and that trust stems from understanding, rather than opacity.

The Long Road to FDA Approval

The path from research to FDA approval took 10 years, including 5 years of algorithm development and validation. What's striking is that Prenosis chose to seek FDA approval even when it wasn't clearly required, recognizing that their tool affects life-or-death decisions.

"We kind of drew a line in the sand and we said, this definitely affects medical decisions. These are life or death situations. And so we want to get FDA approval," Bobby shared.

Remarkably, they remain the only company with FDA approval for AI-based sepsis diagnostics, despite many others working in the space.

Looking Forward: Revolutionizing Drug Development

Bobby's vision extends beyond diagnostics to fundamentally changing how clinical trials are conducted in acute care. By using AI to reveal patients' biological states and subdividing them into subgroups, Prenosis aims to enable precision medicine approaches that could finally make drug development successful in this space.

"What I really hope to show within the next three years is that this type of trial reduces the sample size needed of patients in the trial, but actually increases the signal," Bobby explained.

If successful, this could trigger the kind of pharmaceutical industry attention that acute care desperately needs.

Key Takeaways for Healthcare AI Builders

1. Start with the problem, not the technology: Spend time understanding real healthcare challenges before building solutions. 2. Don't accept data limitations: Consider requiring complete data rather than working around missing information. 3. Build for transparency: Healthcare providers need to understand how AI systems make decisions. 4. Prepare for complex sales cycles: Convincing both clinicians (focused on accuracy) and hospitals (focused on ROI) requires different approaches. 5. Consider regulatory approval early: Even when not required, FDA approval provides crucial validation in the healthcare industry.

The Chicago Advantage

Bobby's story also reinforces Chicago's position as an emerging hub for healthcare AI innovation. His praise for the city's "grounded nature" and hardworking culture reflects the kind of environment where decade-long projects can flourish without the pressure for flashy, quick exits.

As we continue highlighting operators, builders, and thinkers scaling applied AI from the Midwest, Bobby's journey with Prenosis exemplifies the deep, patient work required to solve healthcare's most challenging problems.

The acute care space may have been overlooked for decades, but pioneers like Bobby are proving that with the right approach—combining rigorous science, thoughtful AI design, and relentless focus on real problems—even healthcare's most intractable challenges can yield to innovation.