Concerto Biosciences

A topical that calms eczema by re-balancing skin, not sterilizing it

In a Cambridge lab a few hundred yards from the Charles River, a small team is reading the results of something most of medicine has spent a century avoiding: a treatment that adds bacteria to your skin on purpose. In late 2025, Concerto Biosciences reported that its lead product, a three-strain live biotherapeutic called ENS-002, finished its first human trial in patients with atopic dermatitis. No drug-related adverse events. Eczema symptoms eased. And the troublemaker - Staphylococcus aureus, the bacterium that turns a manageable itch into a flare - was quietly suppressed, while the skin's helpful microbes were left alone.

That last clause is the whole company. Concerto is not in the business of killing things. It is in the business of getting microbes to behave - to perform, as the name suggests, in concert. The skin you are sitting in right now is a crowded ecosystem of competing organisms. Most therapies treat it like a battlefield. Concerto treats it like an orchestra that has lost the tune.

The microbiome is a great idea with a data problem

For two decades, the microbiome has been biology's most promising and most frustrating frontier. We know the bacteria living on and in us shape everything from digestion to skin to mood. We are considerably worse at knowing what to do about it. The standard move - sequence a sample, find a microbe that correlates with health, bottle it - has produced a long shelf of probiotics and a short list of things that reliably work.

The reason is unglamorous: microbes do not act alone. They cooperate, compete, and signal in combinations, and the number of possible combinations is astronomical. Test two microbes together and you have one experiment. Test the interactions across a few dozen species and you are facing millions of pairings. Traditional lab methods test them a handful at a time. At that pace, mapping a microbial community would take longer than the careers of everyone mapping it.

So the founders inverted the question. Instead of hunting for a single hero microbe, they went looking for the relationships - the small groups that, together, push a sick community back toward health. They just needed a way to run millions of those introductions without running out of time, money, or graduate students.

Three scientists, a chip from the Broad, and 18 months of nights and weekends

The technology came first. At MIT and the Broad Institute, Dr. Jared Kehe and colleagues built the kChip - a nanodroplet array that loads microbes into tiny self-assembling droplets, mixes them in every combination, and watches what happens, millions of times over. It was the kind of instrument that turns an impossible amount of biology into a tractable amount of data.

Cheri Ackerman, a Hertz Fellow with a Berkeley chemistry PhD doing her postdoc with Paul Blainey at the Broad, saw what the chip was really for. She, Kehe, and Bernardo Cervantes - later joined by Adil Bahalim - met through the entrepreneurship circuit threading MIT and Harvard. They ran Concerto as a side project for roughly 18 months before fellowships from Activate and a check from The Engine let them go full-time. The company licensed the kChip and incorporated in 2020.

A chip, an AI, and a cast of microbes called an ensemble

Concerto's pipeline runs on two instruments and a vocabulary. The kChip generates the data - structured measurements of how microbes behave in company. A machine-learning layer the company calls kAI learns from those datasets to predict how untested communities will act, so the team can search the combinatorial haystack without testing every straw. The outputs are ensembles: small, defined groups of microbes that, together, do something none of them does alone.

Money that knows the territory

The $23M Series A in November 2022 was led by Safar Partners and joined by Horizons Ventures and M Ventures, the corporate venture arm of Merck KGaA. That last name matters: when a pharma's venture group writes a check, it is also a signal about whether the science could become a product someone wants to partner on. Concerto's earliest support came from The Engine, the tough-tech fund built by MIT, and from Activate's fellowship program - the kind of patient capital that science this early tends to need.

Make microbial communities predictable - and therefore engineerable

Strip away the chip and the chemistry and Concerto's ambition is almost philosophical: take the most chaotic system in biology and make it behave like engineering. If you can predict how a microbial community will respond before you build it, you can design products instead of discovering them by luck. Eczema is the first proof. The company has been clear that the same engine points at vaginal health, consumer skin, and a B2B business where other companies build on Concerto's discoveries.

It is a bet that medicine is shifting from subtraction to addition - from removing the offending cell to restoring the balance around it. Whether that bet pays off at scale is still an open question, and Concerto, to its credit, is answering it the slow way: in a clinic, with endpoints, one trial at a time.

Back in the Cambridge lab

Return to that lab by the Charles. The Phase 1b readout is in, and it says something modest and large at once: you can change a person's skin for the better by adding the right microbes, not by stripping the wrong ones away. Eight patients is not a market. But it is a door, and for a decade the microbiome field has been knocking on it.

Concerto's promise is not a single cream for a single rash. It is a machine for finding ensembles - and an argument that the bacteria capable of healing us are, in many cases, already living on us, just out of tune. If the company is right, the next generation of treatments will look less like an antibiotic and more like a conductor walking up to a disorganized orchestra, raising a baton, and waiting for the room to find the note.