Concurrence
/kənkɜrəns/ noun
1. agreement; combining to produce or bring about something
I started Amigo in 2024 on a simple conviction: everyone deserves good healthcare, and many people will never receive it under the system as it stands. Today, Amigo becomes Concurrence, a name that reflects how we believe the best care becomes available to everyone.
The healthcare system’s problems are familiar to anyone who has interacted with it (long wait times, ever-inflating costs, clinician burnout, poor access, the list goes on), and they have outlasted enough attempts at reform to feel permanent. AI was the biggest opportunity I had seen in my lifetime to address them.
A company represents a hypothesis about the world. The most successful companies are able to ask the right questions at the right time, and each answer determines the universe of questions that become possible to ask next.
Our first question was a practical one.
What kind of care can be delivered by AI?
This question deserves a precise answer because “AI delivering care” can conjure images of robots supplanting doctors to diagnose and prescribe, which is not what we are building. Consequential work like this belongs to humans, and our last two years of deployments have strengthened that conviction.
What we also learned is that diagnostic work, for all its importance, occupies a smaller share of care than most people assume. Surrounding every diagnosis is a vast body of work on which it depends for its outcome: intake, care navigation, scheduling, follow-up, documentation, support between visits, coordination across teams. This work is “care” in the fullest sense of the word, and AI is more than capable of handling it today.
And so, we built AI capable of handling healthcare's most painful workflows, and this first question grew into dozens of deployments supporting more than ten million patients worldwide.
As those deployments matured, something interesting happened. Because our agents were embedded within the care journey, they were capturing new types of data, such as the moment-to-moment reasoning within every interaction and the small, potentially relevant details of patients' lives. What surfaced was a rich record of traces connecting what was done to what happened next. We now call these traces Patient Trajectories, and with enough of them, we can identify patterns that inform how care should be delivered.
As we came to appreciate, moments like this are vanishingly rare in healthcare. What had happened, in effect, was a conversion of knowledge that went unrecorded into evidence an entire organization could act on. Healthcare has never had a mechanism for this, and to understand why, we asked a question so obvious it is easy to miss.
What counts as evidence?
Healthcare prides itself on being an evidence-driven field, but every clinician also carries a second body of evidence—unique knowledge earned over thousands of patients that shapes their decisions at least as much as the literature. No two clinicians have lived through the same cases, which means no two clinicians reason from quite the same evidence base.
I learned what this costs earlier than most. When I was ten, my mom found a lump she was worried about. The first doctor examined her and told her it was nothing. The second did the same. A year passed before a third doctor, alarmed that nothing had been done, ordered tests immediately and diagnosed her with breast cancer. She battled this illness for the next three years before passing away.
I spent a lot of time thinking about those first two doctors. Both examined my mom in good faith and reasoned from the evidence they had—the thousands of patients they had personally seen—but those cases evidently did not inform hers. I wondered if there were doctors somewhere else in the world who had seen a case like hers and would have done things differently.
A staggering amount of information is generated through everyday care delivery, but only a fraction of it is captured. The closest thing we have to shared memory is the chart, which can tell you what happened but is much worse at preserving the full trajectory of care: every question that was asked, which avenues were explored and discarded, why something was tried, and what the implications are for the next similar case. The record we keep is a record of facts, because facts are what a busy clinician can feasibly write down, and billing and compliance are what the writing is for. Everything else evaporates on contact with the next patient that walks into the room.
Worse, the visit is only a fraction of the relevant evidence. Whether care succeeds depends on the mundane particulars of a patient's life, like whether they have transportation to the appointment, the price of the generic drug at their preferred pharmacy, or their preference for text messages over phone calls. These particulars are called social determinants of health, and they are sparsely recorded for posterity.
All of this creates a strange asymmetry. An individual clinician becomes dramatically better over a career by observing thousands of patients, while the organization can treat millions without improving at the same rate. The knowledge exists. It is generated, at enormous expense, every single day. But it accumulates in the heads of individual clinicians rather than in a shared, structured record the whole organization can learn from, and the institution pays for that arrangement in every currency healthcare has—clinician time and patient outcomes above all.
What happens when healthcare learns from every act of care?
This is the question we are now looking to answer, and for the first time in history, we can.
With AI sitting inside the act of care itself, every trajectory can be captured as it unfolds, at a level of detail that after-the-fact documentation cannot achieve, and with no added burden on clinicians. When those trajectories are combined into a shared library (we call this the Health World Model), the next encounter begins from a stronger position than the last.
Ask a hundred clinicians the same hard question today and you will receive a hundred honest answers, each conditioned on a career’s worth of idiosyncratic experience. Grant everyone in the institution the same evidence base, human and AI alike, and their answers begin to converge, because they are finally reasoning from the same priors.
A new name for a bigger question
There is a word for what happens when individuals, given the same evidence, arrive at the same answer. Concurrence means agreement, and specifically the kind of agreement that is earned through empirical evidence.
Renaming a company you love is a hard thing to do. There is an old instinct in company-building that says a name is just a label. I disagree. A name is a claim about what you hope to achieve.
So today, Amigo becomes Concurrence. We chose Amigo for its warmth, and it reflected our bet that AI in healthcare could feel human. That bet paid off across ten million patients, and it will always be a part of our DNA. We chose Concurrence to reflect the scale of what we've built and the question we're asking now.
The terminal goal of technology in healthcare is to make the best care possible available to every human on the planet. We believe the path runs through every act of care learning from the last, and we intend to walk it with the clinicians, patients, and organizations that deliver it everyday.
- Ali