The Intelligence Behind Every Formulation.
This is the substance of Flore Clinical: every formulation is driven by a Central Intelligence Engine built from 9 years of paired microbiome sequencing and symptom-resolution outcomes. The formula didn’t come from a supplier — it came from the data. Here’s exactly how, in plain language. No black box.
What the engine is
The Central Intelligence Engine is a clinical intelligence layer — a structured database of nine years of paired microbiome sequencing and symptom-resolution records, with a query system that an AI uses at the moment of formulation. It is not a product catalog, and it is not a summary of the literature. It is a record of what actually happened to real patients after a specific formulation, and what their microbiome looked like before.
How it was built
Nine years of pairing two things most companies never connect.
Sequencing
14,704 sequencing subjects across 24,193 tests — the microbiome state of each patient, before and, where available, after intervention.
Symptom resolution
What changed for the patient — whether their reported symptoms resolved, persisted, or shifted, captured longitudinally over months.
Pairing those two is the hard part, and it’s the part nobody else has done at this scale. A strain claim from an isolated trial tells you a strain can do something. A paired record tells you what happened when a patient who looked like yours actually took it.
What it does
When you submit a patient’s microbiome data, the engine matches that profile against the outcome patterns of patients who looked similar — and surfaces the formulations that resolved symptoms in those patients. Ingredients are selected because the data says they worked for this kind of patient, not because a supplier promoted them and not because a single paper suggested they might. The result is a formulation matched to the patient profile, grounded in real-world resolution.
Why it’s defensible
Confidence for clinicians comes from transparency, not mystique.
Deterministic
The same patient profile against the same data returns the same logic. It’s a query over records, not an opaque generative guess.
Auditable
Every recommendation traces back to the cohort and the outcome records that informed it. You can ask why a strain is there and get an answer.
Not a black box
AI queries the database; it does not invent the formulation. The intelligence lives in the outcome data, which makes it regulatorily defensible.
The Universal Core
What the data agreed on across every body system.
Three strains appeared in at least 30% of successful formulations across every body system we treat. The data — not a formulator’s preference — elevated them into every adult formula:
Bifidobacterium breve
Lactiplantibacillus plantarum
Lactobacillus rhamnosus
Present in ≥30% of successful formulations across every system.
How you reach it
The intelligence is the product’s engine; the provider portal is how you reach it at the point of care. Through the portal you submit a patient’s microbiome data, order the formulation the engine surfaces, and track whether it resolved their symptoms — then reformulate from the result. The portal is the included access-and-tracking layer, not a separate software product. The therapeutic — the precision formulation, plus GoodOnes™ — is what the intelligence delivers.
Read the methods
The full white paper details how targeted, outcome-derived probiotics are formulated and validated.
Read the white paper → See the evidence →