AI Solopreneur TrendDiscover the World's AI Solopreneurs

Oct 20, 2026 · nursingworld.org

A Nurse Who Was Dismissed as a Patient Built an AI Advocacy App on Her Own: Rachell Dumas and HEARD

After years as a critical care nurse and health informatics specialist, Rachell Dumas was sent home from an emergency room with stroke symptoms. She turned that experience into HEARD, an AI-powered app that helps patients document symptoms and speak up, and she runs it as a solo founder with an advisory board.

Nurses see, every shift, where patients get lost in the healthcare system. Few of them get to build the fix. Rachell Dumas did, after the system failed her as a patient.

This article is for nurses and other healthcare professionals who have an idea for a tool and wonder whether one person can build it. We look at what she built, how she runs it alone, and then, in a separate section, at what her nursing experience brings that a software team would not.

Who she is

Rachell Dumas, MSN, RN, is a registered nurse with more than a decade of clinical experience, including trauma, neuro ICU and COVID care. She also spent more than a decade implementing and maintaining electronic health record systems, holds an MSN in Nursing Informatics and is a doctoral candidate in the same field.

The experience behind the product

According to the American Nurses Association's podcast SEE YOU NOW, Dumas went through nine pregnancy losses over four years before her son was born. Six months later, she woke up unable to see. At the emergency department, with vision loss and stroke symptoms, she was sent home with an antacid. As a neuro ICU nurse, she knew something was seriously wrong, went to a second emergency department and was eventually diagnosed with idiopathic intracranial hypertension. Treating it took two brain surgeries.

On HEARD's website she puts it this way: "I almost lost my life because my voice didn't matter."

What she built

HEARD stands for Human Enabled AI Reporting and Documentation. It is an AI-powered patient advocacy app for patients and caregivers.

  • Symptom documentation. Patients keep a dated record of symptoms, with photos, voice notes and timestamps, instead of describing months of problems from memory in a short visit.
  • Appointment preparation. The app helps patients rank their concerns and prepare the questions to ask.
  • Escalation. When a concern is brushed off, HEARD helps patients write a clear message to their provider or hospital administration.
  • Caregiver coordination. Family members work from one shared timeline, so the story does not have to be repeated to every new clinician.
  • Plain-language support. According to the ANA podcast, the platform also translates medical jargon and connects users to second opinions and legal support when needed.

HEARD states on its site that it does not provide medical diagnoses or medical advice.

How she runs it alone

HEARD's team page lists one person in an operating role: Dumas herself, as founder and CEO. Everyone else on the page is an advisor, including a lawyer, a reproductive endocrinologist, a health scientist, a certified patient advocate and a health tech investor.

  • Product stage. Patients can use HEARD free during an early access period. The web app and an interactive demo run on Lovable, an AI app builder.
  • Two kinds of customers. Besides patients and caregivers, HEARD offers employers, benefits brokers and health systems a version with aggregated, de-identified "Benefits Intelligence" reports on how people use their benefits and where they get stuck.
  • Telling the story herself. She has told HEARD's story on nursing podcasts, including the ANA's SEE YOU NOW and Nurse.org's Club Nurse, and on stages such as Atlanta Tech Week.
  • Next steps. The site has a page for investors, which suggests she may raise money and build a team later. For now, the operating side is hers alone.

What nursing experience gave her that a software team would not

This section is our analysis, not a claim made by Dumas.

  • Knowing where patients get dismissed. Each feature maps to a moment she knows from both sides of the bed: the rushed visit, the forgotten question, the concern that goes nowhere.
  • Knowing how records work. Her years in health informatics show in the product's focus on documentation, timelines and structured records, which is how clinicians and hospitals actually process information.
  • Knowing the limits. As a nurse, she draws a clear line: HEARD helps patients organize and communicate, and leaves diagnosis and treatment to clinicians.
  • A story that earns trust. In healthcare, users and partners need a reason to trust a new tool. Her experience as both nurse and patient is that reason, and the nursing community has amplified it.

What you can copy

These are our suggestions, drawn from her story. They are not medical or legal advice.

  1. Start from a moment you know well. Her product begins with a specific failure she lived through, not a broad idea about "fixing healthcare."
  2. Use AI to organize, not to diagnose. Documentation, preparation and communication are useful places for AI and avoid making clinical decisions.
  3. Build a prototype people can try. A working demo on an AI app builder lets patients, nurses and partners react before you raise money or hire.
  4. Borrow expertise through advisors. An advisory board covering law, medicine and patient advocacy fills gaps without adding employees.
  5. Tell the story in your professional community. Nursing podcasts and associations are a natural audience for a nurse-built tool.

Notes on the sources

The account of Dumas's medical experience comes from the American Nurses Association's SEE YOU NOW podcast page, published on April 29, 2026; her clinical and informatics background also comes from a Nurse.org article published on April 2, 2026 and from HEARD's website. Details about the product, the team and the business come from HEARD's own website and are as stated there as of October 2026. HEARD's membership page lists plans at $15, $25 and $40 a month, which take effect at launch; use is free during early access.

Sources

Related articles