Understanding Maggie: What our AI assistant can and can't do

Dayhoff Technologies Inc. ("Dayhoff Health")

Last updated: July 30, 2026

Maggie is the AI assistant built into your Dayhoff account. She's there to help you make sense of your kit results — reports that are dense, technical, and unfamiliar to most people the first time they see one. Maggie can walk you through what a particular marker or measurement is, how your sample was analyzed, what the numbers on the page represent, and what the current scientific literature says about them.

This page explains what Maggie is, what she isn't, and why you should treat everything she says as a starting point for a conversation with your healthcare provider rather than a conclusion about your health.

Maggie is not a healthcare provider

Maggie is a software tool built on a large language model. She is not a doctor, physician assistant, nurse practitioner, registered dietitian, nutritionist, pharmacist, or any other licensed clinician. She has no license to practice medicine in any jurisdiction, she is not supervised by a clinician in real time, and no doctor-patient relationship is created by using her.

That distinction matters more than it might seem. A clinician who sees you brings things Maggie fundamentally cannot: your full medical history, your medication list and the interactions within it, your family history, your physical exam, your other lab work, your symptoms as you actually experience them, and years of training in weighing all of that against a single test result. Maggie has your report and whatever you type into the chat window. That is a very narrow slice of you.

Specifically, Maggie does not and cannot:

• Diagnose anything. She cannot tell you whether you have a condition, disease, infection, deficiency, or disorder — including ones your report might seem to point toward.
• Recommend treatment. She cannot tell you what to take, what to stop taking, what dose to use, or what protocol to follow. This includes prescription medications, over-the-counter products, probiotics, prebiotics, supplements, and herbal preparations.
• Prescribe a diet. She can explain what research says about, for example, dietary fiber and microbial diversity. She cannot design an eating plan for you, tell you to eliminate a food group, or set targets for you to hit.
• Order or interpret additional testing. She cannot tell you which follow-up tests you need or read results from tests we didn't perform.
• Assess urgency. She cannot tell you whether a symptom is serious, whether something can wait, or whether you should go to an emergency room.
• Replace, second-guess, or overrule your provider. If Maggie says something that conflicts with what your doctor told you, your doctor is the one who knows your case.

Maggie can be wrong — sometimes confidently

Large language models like the one behind Maggie generate text by predicting what should come next based on patterns in their training data. This works remarkably well much of the time, and it also means the system has no internal sense of when it has crossed from "recalling something accurate" into "producing something plausible-sounding but false." The industry term for this is hallucination.

In practice, this means Maggie may:

• State a fact about an organism, molecule, or biological pathway that is outdated or simply incorrect
• Cite a study, author, journal, or statistic that doesn't exist, or attribute a real finding to the wrong source
• Present a contested or preliminary research finding as though it were settled science
• Overstate how strong the link is between a measurement in your report and a health outcome
• Misread a number, unit, or reference range in your own report
• Give you a different answer than she gave last week, or than she gives another customer with a similar result
• Miss something in your report that a trained eye would catch

Errors can also compound. If Maggie gets a detail wrong early in a conversation, the rest of that conversation may be built on top of the mistake. She writes fluently and confidently regardless of whether she's right, so fluency is not a signal you can use to judge accuracy.