GutMetabAI simulates how a food or supplement shifts a person's gut microbial community, and turns that into a personalized recommendation grounded in metabolism rather than a generic guess. It reasons from the biochemistry of your own microbes. A working prototype today.
The same food helps one person and does nothing for another. Whether it works often depends on whether your gut microbes carry the enzymes to act on it, and generic advice, and generic AI, ignore that entirely. The result is confident recommendations that quietly fail for most people.
A mechanistic simulation of your microbial community. Give it your microbiome and a candidate food, and it predicts how the community's metabolism shifts, which foods are worth trying, and, for each one, the molecular reason. The numbers come from a metabolic model, not a language model.
The same grounded core as the rest of biotech: deterministic engines do the science, the model reasons from biochemistry.
Each species in your sample gets its own genome-scale metabolic model, so the simulation reasons from what those microbes can actually do, not from correlation.
Community flux-balance analysis simulates how those microbes feed each other and compete on a given diet. It is deterministic math over the whole community, not a guess.
Change the diet and re-solve. The model predicts how the community and the metabolites it produces shift when you add a specific food or supplement.
Before it recommends a food, it checks whether your microbiome even carries the enzymes to activate its compounds. Many people lack the microbes to convert a given nutrient, so a food that helps on average can do nothing for you. This is where the personalization lives.
GutMetabAI is a working prototype, not a shipped product. The pipeline runs real community flux-balance simulations locally today; it is not publicly available yet. It is built for regional functional foods and gut microbiomes that differ from the Western guts most nutrition tools are tuned on. If this is your problem space, tell us below.
Clinicians, food and supplement teams, and researchers: tell us what you are trying to answer, and we will show you what the simulation can do.