By Craig Rouskey, CEO and microbiome scientist, Flore Inc.
In the last post I made the case that you're a doughnut — that your gut is the interface where the outside world negotiates with the inside of you. This post is about what that interface can actually do, and why I think it's one of the most underused pieces of biotechnology we already own.
Here's the reframe: your gut is a bioreactor.
An absurdly good system that nobody calls a bioreactor
Industrial bioreactors are giant stainless-steel tanks where companies grow microbes to produce things — insulin, enzymes, biofuels, vaccine components, niche pharmaceutical intermediates. You pick a strain, you feed it the right substrate, you control the temperature and pH, and you harvest what comes out. It's how a huge fraction of modern medicine gets manufactured.
Your gut runs the same operation. Body-temperature regulated. pH-stratified along its length. Continuous substrate flow three times a day. Trillions of microbes across hundreds of species, each one a metabolic specialist. The output gets absorbed directly through a single-cell-thick membrane into the bloodstream and routed to every tissue in your body. From a chemical engineering standpoint, it's an absurdly good system. We just don't usually think of it that way because it's quiet and it doesn't have a logo.
In principle, you can use this bioreactor to produce almost any small molecule you want and deliver it directly to the body — no pill, no injection, no first-pass liver metabolism stripping out 80% of the dose before it ever does anything.
The molecule gets made in situ, on the outside of the membrane, and walks across the interface continuously. Steady-state dosing without dosing.
This isn't theoretical — it's already happening, accidentally
Gut microbes decarboxylate tyrosine into tyramine. They convert tryptophan into a whole catalogue of indoles and signaling molecules. They produce short-chain fatty acids that affect everything from colonic energy supply to systemic inflammation. Methanobrevibacter smithii pulls hydrogen out of the system and shifts the metabolic economics of every other organism around it. The bioreactor is running. The question is whether you know what it's making, and whether what it's making is what you'd choose.
This is where the data starts to matter
This is where Flore's data starts to matter. Nine years and 23,447 patients of longitudinal microbiome taxonomy gives you something most people working in this space don't have: a map of which strain combinations correlate with which metabolic outputs, in which kinds of people, over time. You can start to ask questions like:
- Which Firmicutes consortium is doing the decarboxylase work in this person?
- What's the AADC gene picture at strain resolution?
- If we want more of compound X showing up at the gut wall, which community structure produces it — and how do we shift toward that structure?
What the interventions actually look like
The interventions that come out of this framing don't look like traditional drugs. They look like targeted ecological edits. You're not delivering a molecule — you're delivering an organism, or a substrate, or a removal of something that's been suppressing the organism you want. The molecule comes from the bioreactor, not from a factory in New Jersey.
The downstream possibilities are wide:
- Mood compounds without psychiatric pharmacology.
- Anti-inflammatory output without immunosuppression.
- Anxiolytic precursors produced continuously instead of dosed twice a day.
The next target I'm thinking about — Prevotella copri in the context of anxiety — sits exactly in this frame: not a drug for anxiety, but a community shift that changes what the bioreactor outputs, and lets the rest of the body respond to a different chemical environment.
A pill is one thing. A bioreactor is another.
A pill is a one-shot delivery problem. A bioreactor is a continuous manufacturing problem. They are not the same thing, and we've been treating the gut as the former when it's clearly the latter.
We already own the factory. Time to start reading the production line and adjusting it on purpose.
Frequently asked
What does it mean to call the gut a bioreactor?
It runs the same biochemistry as an industrial bioreactor — body temperature, pH stratified, continuous substrate flow, trillions of microbial cells producing molecules that are absorbed directly through the gut wall into the bloodstream. The bioreactor framing is engineering, not metaphor.
What kinds of molecules can the gut bioreactor produce?
Short-chain fatty acids (acetate, butyrate, propionate), indoles, tyramine, B-vitamins, bile-acid metabolites, neurotransmitter precursors, anxiolytic precursors — gut microbes are already producing these and more. The point is that we can direct what gets produced.
Why is in-situ production better than a pill?
A pill loses most of its dose to first-pass liver metabolism. A molecule produced at the gut wall and absorbed continuously avoids that loss, delivers steady-state concentrations, and doesn't require repeat dosing. It's continuous manufacturing instead of one-shot delivery.
What is Flore doing with this idea?
Using nine years of longitudinal microbiome data across 23,447 patients to map which strain consortia produce which metabolic outputs, then designing personalized probiotic formulas as targeted ecological edits — delivering organisms, substrates, or removing suppressors so the bioreactor outputs what each individual needs more of.
Read previous: You're a Doughnut — why gut health is about the barrier, not the bowels.