It begins quietly, in a room that smells of hand sanitizer and old wood, where a certified nursing aide barely has time to glance at the clipboard before starting a shift. The chart says “Room 214, medication at 6 p.m., assistance with mobility, dietary restrictions.” It does not say that the woman in the bed once danced on a stage in Chicago, that she met her husband at a USO dance in 1952, or that she still hums Glenn Miller songs when she thinks no one is listening. It does not say that the man across the hall spent thirty years as a machinist, that his hands built parts for airplanes, that he taught his daughter how to fish before she could tie her own shoes. In the crush of daily care—medications, transfers, meals, documentation—the human story is the first thing to get lost. This is the problem that Patrick Donahue, a Seattle-area software engineer, could not stop thinking about. The son of medical professionals, Donahue grew up with the quiet presence of hospice care in his family life. He knew what it meant to be present for someone at the end of their life, and he knew how much of a person’s history can slip away if no one thinks to ask the right questions while there is still time. When his own father died, that knowing became visceral. He had kept a journal for as long as he could remember, and he knew what putting a life into words could do—for the person speaking and for the people listening. From that grief and that habit of reflection, an idea was born: a small, gentle piece of technology that could help residents in assisted living and memory care communities tell their own stories, and make sure that the people caring for them actually knew who they were. He called it Porchlight.
Porchlight is not another companion chatbot, not a screen with a smiling avatar, not a surveillance device disguised as wellness technology. It is simpler, and in some ways more radical. A resident taps a single button on a device and hears a question—about a first job, a favorite holiday, a person who changed their life. They talk for as long as they want. The platform records their voice, then uses AI transcription and language models to transform that raw, rambling, beautiful mess of memory into two things. One is a “Know Your Resident” briefing, a concise one-page summary for care staff that captures what matters: the people they loved, the work that shaped them, the small joys that can still light up a face. The other is a private audio feed for family members, a place where children and grandchildren can listen to their mother or grandfather tell a story in their own voice, and can reply with their own voices, creating a conversation across time and distance. Donahue, who is the founder and sole employee of Porchlight, launched the tool in 2026 after a year of building and listening. He has no paying customers yet, and he has had fewer conversations with activity directors than he would like—but the conversations he has had taught him something crucial. The first question was never about the technology. It was, “What does it cost?” and “Who has time to set it up?” That pushed him to make setup almost laughably simple: a printed QR card, placed on a bedside table or a wall, that links directly to the experience. No training manual, no login dance, no IT ticket. And because he is asking communities to trust him with their residents’ most intimate memories, he is making it free while they try it. The scale of his ambition is not to build a sprawling startup teeming with managers and roadmaps. It is to get into enough rooms, enough lives, that the culture of care changes in one meaningful way: the aide who walks in for the first time knows more about the person in the bed than a diagnosis and a room number.
Donahue’s obsession is rooted in his own history, and he speaks about it with a gentleness that suggests he has carried this grief for a long time. Both of his parents worked in medicine, and hospice care was not an abstract concept but a fixture of family life. He saw how a good death depends on being known, on having people who understand your history, your humor, your fears. When his father died, he realized how much of a person’s story goes with them unless someone asks while there is still time to ask. He had always believed in the power of writing down your life—his journal was proof of that—but he also knew that not everyone is a writer, and not every family has the language or the courage to ask the big questions on their own. Porchlight was the tool he wished had existed when he was sitting at his father’s bedside, trying to remember everything, trying to hold on to the details that would be lost. The design is deeply influenced by people who understand memory and meaning. Donahue has been reading John Zeisel’s “I’m Still Here,” a book about connecting with people living with dementia, and listening to the StoryCorps podcast, the extraordinary archive of ordinary people telling their own stories. His entrepreneurial idol is Dave Isay, the founder of StoryCorps, because Isay proved that the simple act of asking a person to tell their story, and recording it faithfully, is worth building an entire institution around. Donahue is not trying to build an institution, not yet. But he is trying to build a practice, a way of weaving storytelling into the fabric of daily care so that it does not feel like another task on a checklist but like the best part of the job—the part where a resident’s eyes light up because someone finally asked about the dance where he met his wife. This is the quiet revolution at the center of Porchlight: not better analytics, not more efficient workflows, but the restoration of personhood in places where personhood is often lost in the grind of routine.
Artificial intelligence is the reason Porchlight can exist at all, but Donahue is careful to keep it in its place. AI is not the product; it is the quiet background worker that makes human connection possible. Transcription software has become good enough to understand older voices, including the softened consonants of a stroke survivor or the accent of someone who grew up in another country. Language models can take an hour of wandering memories—a story about a dog, a detour into a recipe, a long pause, a laugh—and distill it into a one-page briefing that a caregiver can read in ninety seconds. None of this replaces anything human; it simply removes the friction between a resident’s story and the person who should hear it. Donahue thinks of the impact in three directions. First, with family: the family relay lets a resident’s children and grandchildren hear their stories in their own voice and reply with their own voices, creating an heirloom that is not frozen in amber but alive, relational, ongoing. Second, with neighbors and community: as residents record their memories, the platform can help turn those individual stories into community events, newsletters, and shared moments—a resident telling the story of growing up in a small town, and another resident recognizing a place they once visited. Third, with the care team: staff know the person, not just the chart. An aide can say, “I heard you used to play piano,” and the conversation changes. A resident who was disengaged suddenly has a bridge. A person who was labeled difficult is revealed to have been a soldier, a teacher, a mother of five. The technology is not pretending to be a person. It never offers advice, never tries to be a friend. It simply listens, organizes, and returns the story to the people who should have had it all along. On the build side, AI has changed what is possible for a bootstrapped solo founder. Donahue has spent ten years as a software engineer helping startups scale, and he credits AI coding tools with giving a one-person company the speed and polish that once required a team. He can build, test, and refine features in days instead of months. But he remains adamant that the goal is not efficiency for its own sake. It is a better quality of life, more dignity, and a way for residents to pass their wisdom on to the people they love. The resident talks. Real people listen. The machine is just the bridge between them.
There is an honesty in Porchlight’s early path that is both admirable and rare in startup culture, where hype often outpaces reality. Donahue made the hardest ethical choice of his young company this past year by pulling claims off his own website. When he checked his marketing language against the actual code, he found that some of what he was describing were planned features, not built features. The words made the product look more complete than it was, and that was exactly the problem. He was asking communities to trust him with their residents’ stories, and trust, once spent, is almost impossible to earn back. So he cut the claims, even though it made the product look smaller at the exact moment when he wanted it to look bigger. It was a quiet act of integrity, and it defines the way he wants to build. His advice to other entrepreneurs is rooted in that experience: make sure what you say matches what you have built. It is tempting to describe the version of the product that lives in your head, especially when you are pitching investors or trying to win early users. But early on, trust is most of what you have to sell, and it is easy to spend it without noticing. Donahue is not immune to the pressures of being a solo founder. He wants Porchlight to succeed, wants communities to adopt it, wants to prove that a tiny bootstrapped company can make a difference in a sector that is often overlooked by Silicon Valley. But he is willing to go slowly, to have honest conversations, to accept that he does not yet have paying customers and needs to earn them one conversation at a time. He has had fewer customer conversations than he wants, and he does not pretend otherwise. He is learning, adjusting, and letting the product be shaped by the real needs of activity directors and care managers, not by his own assumptions. That humility is part of the product’s design. It is a listening tool, after all—and the person who built it is also learning to listen.
What would success look like for Porchlight? Donahue describes it with a moment, not a metric. A new aide walks into a room on the first day, and instead of just checking a vitals monitor, she says, “I heard you met your wife at a dance. What was that like?” And the resident, who has been mostly unresponsive all morning, lifts his head and begins to tell the story. His eyes brighten. His voice gains rhythm. For four minutes, he is not a patient. He is a young man at a dance, nineteen years old, heart pounding, watching a girl across the room. The aide does not rush him. She listens. And later, the community pays for Porchlight because that moment is worth something to them—not just as a nice thing that sometimes happens, but as a reliable, repeatable part of how they care for people. The vision is larger than the tool itself. It is about the end of loneliness in elder care, the recognition that a person is not their diagnosis, the belief that every single human being has a story that matters to someone. It is about the daughter who never knew her father had been a ballplayer until she hears him say it in his own voice, three years after she thought she knew everything there was to know about him. It is about the granddaughter who replies with her own story, and the two of them, separated by age and memory, meeting in a shared moment of truth. It is about dignity, legacy, wisdom, and the simple, profound act of asking someone about their life. The porch light is a metaphor for welcome, for the warm glow that tells a person someone is home, someone is listening. In a world where so much technology is built to automate away human interaction, Porchlight is built to amplify it. It will not replace caregivers. It will give them a reason to stay in the room a moment longer, a question to ask, a story to receive. That is the future Donahue is building, one recorded memory at a time. And if he is right, the first thing a caregiver sees when they walk into a room will not be a chart. It will be a person.
