Contact

Technology

Technology

A complete system for discovering and validating skin mechanisms and bioactives—from computational prediction through human tissue testing.

  1. 01YUNA: Long-lasting human tissue

A full stack of capabilities under the same roof, working as one loop–each compound moves from in silico to human-relevant data, and everything we learn improves the next prediction.

01

YUNA: Long-lasting human tissue

Capturing the full complexity of human skin–because it is human skin.

We work with full-thickness human tissue — epidermis, dermis, multiple resident cell types, and preserved immune-associated molecular programs.

Ex Vivo Tissue Lifespan

Days

1
10
20
30

  • Conventional ex vivo skin: ~1 week

  • Outer Bio: 30+ days

Extended viability enables research beyond conventional short-duration skin models

  • Chronic inflammation studies

    Sustain chronic inflammatory responses over multi-week windows and evaluate whether compounds suppress or modulate inflammation-related pathways

  • Senescence and aging studies

    Run 3-week protocols on older and diverse donor tissue to capture aging-associated responses over time

  • Collagen and matrix remodeling

    Track matrix remodeling, collagen synthesis, and tissue-level changes over weeks

  • Long-term compound and product testing

    Evaluate compounds and products across broader skin-health targets using longitudinal, multi-modal readouts

A tissue supply chain built to capture human diversity

Skin is one of the body’s most heterogeneous organs. Many standard models flatten that variation; ours is built to capture it.

We source full-thickness skin from donors across diverse donor populations and incorporate donor diversity directly into our research and discovery process. This helps us look beyond a single donor profile to identify responses that are broadly shared across human skin, as well as responses that may vary across donor groups.

How we source samples

Our platform uses

All gendersAll Fitzpatrick Types (I-VI)Multiple body regionsMultiple racesAges spanning 20-75
02

Real-World Stress Testing

Damage happens in the real world. It can happen in our research, too.

YUNA is an open-top system, so we can expose skin to the environmental stressors that drive real-world damage — and measure the response over meaningful timeframes.

  • UV Damage

    Calibrated UVA/UVB exposure; track damage accumulation, repair kinetics, and protective interventions over days to weeks

  • Barrier disruption and dryness

    Tape stripping, chemical exposure, or low-humidity stress to test barrier repair agents on compromised skin

  • Allergen-relevant and irritant challenge

    Common sensitizers and irritants; monitor skin stress responses and recovery

  • Wound and mechanical injury

    Controlled incisions or punch biopsies for re-epithelialization, matrix remodeling, and wound repair-associated responses

03

Automated phenotype screening

Between prediction and human tissue, we rapidly validate predicted hits on focused in vitro models of the most important phenotypes in skin health. Only the strongest candidates advance to real skin—and every screen generates large amounts of training data along the way.

Focus areas

Anti-oxidantAnti-inflammatoryCollagen SynthesisSenescenseSkin repair/rejuvenationWound healingToxicity
Our research

  • Robotic liquid handling

  • Automated microscope

04

ML prediction & in silico screening

Scalable screening made possible by machine learning

Before we test a compound in tissue, we use machine learning to predict whether it's likely to be active. Our predictions are trained on data from our own platform — not just public datasets. The models improve with every round of testing.

CURRENT CAPABILITIES

  • 750M compounds per year

    Our foundation model pre-screens vast chemical space, surfacing novel, high-potential actives before any are synthesized.

  • 30,000+ data points per compound

    Every compound we test on tissue generates deep transcriptomic, biochemical, and histological signals—the data that trains our models to recognize what truly drives human skin health.

  • Machine learning for efficacy, safety, IP and manufacturability

    We score and de-risk every compound across all four dimensions, rapidly surfacing the most viable candidates.

Novel target discovery

Most skincare optimizes against a short list of known endpoints. Our platform does the opposite: it surfaces new biological targets from our proprietary human-tissue data, then ranks the chemical structures most likely to engage them. We're not reformulating old mechanisms — we're finding new ones.

05

A self-reinforcing learning loop

A platform that gets smarter with use

Every experiment adds to a dataset that improves the next prediction. The compounds we test today make tomorrow's models more accurate—refining which targets to pursue, which compounds to prioritize, and where to look next.

  • Mechanism identification

    Understand how a compound works, not just that it works

  • Biomarker development

    Identify molecular signatures that inform clinical response

  • Faster iteration

    Predictive models improve with each experiment, reducing time to validated hits

Molecular data learning loop

We map our findings to phenotypic skin function to create a learning loop where molecular signatures predict functional outcomes, guiding discovery toward compounds that will actually work in real skin.

Experiment

Molecular Data

Phenotypic Skin Function Mapping

Functional Outcome Prediction

Learn more about our research

Interested in working with us?

We partner with beauty brands, ingredient companies, and pharmaceutical developers.

Let’s start a conversation.