The Palisades fire saved my life. Not metaphorically—literally. When flames consumed our home in January 2025, we fled to Palm Desert. A couple weeks later, severe abdominal pain struck during a steak dinner with friends landed me in an unfamiliar Rancho Mirage ER. Those doctors, unburdened by months of “inconclusive” tests and “fishing expeditions,” saw immediately what everyone else had missed: a rare blood cancer ravaging my body. Rogue plasma cells were producing toxic misfolded proteins instead of antibodies. The measure of these Kappa light chains, which should be under 19.4, had climbed to 487.7 mg/L – 25 times normal – or high enough to destroy my organs if left unchecked.
For months before the fire, I’d known something was wrong. At 60, I’d lost weight, my energy had drained, my appetite had vanished. I pushed my doctors for every test imaginable – scans, scopes, labs, cardiac assessments. There were plenty of abnormal results, but nothing that stood out to my doctors. They suggested stress, gas, maybe depression.
During my nine-day hospital stay in Rancho Mirage, one question haunted me: My new doctors had reached this diagnosis but why hadn’t my earlier doctors seen it in the same data?
I had some advantages in exploring this question. I’d spent decades working in healthcare technology—founding Health Hero Network, pioneering remote patient monitoring, earning over 200 patents in health tech and AI. I understood how medical systems worked, and more importantly, how they failed. Now as the patient, I knew exactly what tool I needed.
The Diagnosis Hidden in Plain Sight
I exported my full MyChart record – labs, scans, doctor notes from months of ambiguous tests – and fed it into an AI agent I named Haley. Powered by models from OpenAI, Google, Anthropic and xAI, Haley was layered with structured medical reasoning.
I watched her process my labs in real-time. “Mild anemia present,” she noted. “Combined with low immunoglobulins and elevated ferritin, this pattern suggests possible plasma cell dysfunction. Recommend serum free light chain test and bone marrow biopsy.” The tests Haley recommended had never been suggested during those earlier months.
Encouraged, I spun up more agents: an oncologist, hematologist, nephrologist. Each reviewed my data independently. Then I created a synthesis agent—Hippocrates—to analyze their findings and issue unified recommendations.
The consensus was clear: “Bone marrow dysfunction. Test immediately.”
My new doctors had found this—but AI showed it had been findable all along, hiding in the patterns of my earlier test results.
From Diagnosis to Precision Treatment
But diagnosis was only half the battle. On February 14, I started the standard treatment: Daratumumab + CyBorD. It’s what guidelines recommend as first-line therapy. But when I fed my cancer’s cytogenetic profile into my AI agents, they flagged something crucial: my cancer cells carried a t(11;14) translocation that suggested superior response to venetoclax.
Venetoclax for plasma cell cancers like mine hasn’t had a full randomized controlled trial as first-line therapy—it’s used off-label by some leading cancer centers based on compelling biological evidence. AI models have synthesized outcomes from patients who received venetoclax at various points in treatment, compared cytogenetic profiles and predicted response rates.
With my Kappa levels at 487.7—threatening irreversible organ damage—I didn’t have time for trial and error. I needed the best treatment first. Failure wasn’t an option.
I consulted specialists at UCSD, UCSF and Mayo Clinic. I walked in with mathematical models showing that my predicted resistance to standard care was already happening and evidence that venetoclax matched my cancer’s specific genetic variation. The conversations weren’t about asking permission, they were collaborative discussions about optimizing my treatment based on my cancer’s unique profile.
My specialists agreed. In April, I switched to Daratumumab + venetoclax as frontline treatment. By August 8, my Kappa levels had plummeted to 14.0 mg/L—a deep response that standard treatment likely wouldn’t have achieved. We’d prevented the organ damage that could have led to heart or kidney transplants.
Building CureWise: AI Knowledge for Every Patient
Every month when I get my IVIG infusion—antibodies from healthy donors keeping me safe while my immune system rebuilds—I think about others facing similar battles without access to these insights. How many patients aren’t getting the best possible treatment given their unique cancer genetics and what’s known today? How many are following standard protocols when their cancer’s specific mutations point to better options?
This is why I’m building CureWise—driven by both mission and the belief that democratizing medical intelligence can transform cancer care. While concierge medicine remains out of reach for most people, CureWise makes advanced medical insights accessible.
Users upload their medical records and labs and CureWise lets them talk to their medical records—to understand what they mean, what treatments might be available and what they should discuss with their doctors. It’s about learning about your disease so you can advocate for the best care available. The system that helped me understand my t(11;14) translocation’s significance can help others understand their unique medical situations.
The Future of Precision Oncology
What this journey taught me is that AI finds subtle patterns that people can miss and never assumes symptoms are “probably stress.” When used responsibly, AI doesn’t replace physicians, instead it empowers both doctors and the patient with insights that would otherwise remain buried in data.
Cancer will claim over 600,000 Americans again this year. It remains very much an unsolved problem of staggering complexity. But precision oncology—getting the most targeted treatment possible for each unique disease—is where we’ll find answers. Every cancer is different, with its own genetic fingerprint, its own vulnerabilities, its own optimal treatment path.
Today’s AI foundation models contain vast troves of medical knowledge, but they’re black boxes—it’s hard to know how they reach their conclusions or whether to trust them. That’s why I’m building swarms of specialized agents that examine medical questions from multiple perspectives—an oncologist agent, a genetics counselor agent, a clinical trials specialist agent—then synthesize their findings. This multi-agent approach makes AI’s medical insights more transparent and trustworthy, turning opaque knowledge into actionable intelligence.
CureWise exists to bridge this gap—to make complex medical knowledge accessible to every patient. Because no one should have to build their own AI system to understand their disease. No one should suffer through suboptimal treatment when their cancer’s genetic fingerprint points to better options. No one should lose precious time to diagnostic confusion when patterns exist in their data.
My wife had already lost our home to the Palisades fire. The thought of her losing me too drove me to find answers. But the tools I built to save myself shouldn’t be mine alone. They should be available to anyone facing a complex medical journey.
Today, my Kappa levels hold steady in remission range. My next bone marrow biopsy will confirm what the numbers suggest: we found my cancer’s lock and used the right key from the start. That’s the future CureWise is building—where every patient can understand their unique disease, where medical knowledge becomes accessible rather than arcane, where AI transforms patients from passive recipients to informed advocates.
The system that helped save my life will soon be ready to help others understand theirs. Because in the end, this isn’t about replacing doctors or promising miracle cures—it’s about ensuring every patient has the knowledge and tools to advocate for their best possible care.
Steve Brown is the founder and CEO of CureWise, a health technology company dedicated to democratizing precision oncology insights. A serial innovator in healthcare technology, he previously founded Health Hero Network, pioneered remote patient monitoring, and holds more than 200 patents in health tech and AI. His personal battle with a rare blood cancer inspired the creation of CureWise, giving every patient access to the advanced medical intelligence that helped save his life.
