We describe what MedFlow does for a practice in three words: See It, Close It, Prove It. That is the value, and it is written from the practice owner's side of the desk.
Underneath it there is a second set of three words that describes what the engine actually does: find, automate, validate. That is the mechanism, and the order is not arbitrary.
Human first. Business second. Proof third.
Find: human first
The loop starts with a person, not a rule. The engine reads the record continuously and identifies the patients who are due for something: a screening, a follow-up, a lab that was ordered and never resulted, a referral that left the building and never came back.
This is the step most often mistaken for reporting. A report tells you a list exists. Finding, in the sense we mean it, means the patient is surfaced to the people who can act, at the moment they can act, inside the system they are already working in.
If this step is wrong, everything downstream is wasted effort applied to the wrong panel.
Automate: business second
Once the patient is found, something has to happen. The outreach has to go out. The order has to be placed. The referral has to be routed and then chased until it closes.
This is where most automation projects stop being about technology and start being about operations. The happy path is easy. What matters is the exception: the patient who does not answer, the referral the specialist bounces back, the order that fails a payer rule.
Automating the happy path and leaving every exception to a person does not reduce the work. It concentrates it, and the people left holding it are the ones who were already stretched.
Validate: proof third
The last step is the one most often skipped, and it is the reason programs survive or die.
Every action the engine takes is recorded against the record it touched, as it happens. That means the completion rate is a fact the practice can read at any time, rather than a story reconstructed at the end of a quarter from data that was never kept for the purpose.
Proof is not a reporting feature. It is what lets a practice defend a program when somebody asks whether it is worth continuing, and it is what makes a value-based contract settle on evidence rather than assertion.
Why the order matters
Reverse any two of these and the loop stops working.
Automate before you find, and you have built an efficient way to work the wrong list. Validate before you automate, and you are measuring a process that still depends on somebody remembering. Find and automate without validating, and the program does real good for a year and then gets cut anyway, because nobody could show what it did.
Find, automate, validate. That is the whole engine, and every care pathway we run is the same three steps applied to a different clinical question.