BNA AI-PLATFORM

We map how botanical molecules interact with human receptors.

NVIDIA Inception Program

WHAT BNA™ IS

Physical AI for aroma and flavor.

Botanical Neuromodulating Assembler (BNA™) is Eybna’s AI platform for the design, compliance, production and logistics of functional aroma.

You describe what you want in plain language. BNA composes the candidate formula, predicts how it will smell and taste, clears it against the rules of your market, checks what is fresh on the shelf, and opens the work order in the lab.

The output is a bottle. That is what physical AI means: the model starts with a prompt and ends with an ingredient we manufacture and ship to your team.

NVIDIA Inception Program
Non GMO seal

FSSC 22000

Non GMO seal

ISO 9001

Food Grade seal

Food Grade

Only Natural Ingredients seal

Only Natural Ingredients

Solvent Free seal

Solvent Free

Botanical Derived seal

Botanical Derived

OU Kosher seal

OU Kosher

Non GMO seal

Non GMO

Gloved hand examining a cannabis flower and an extract vial with a magnifier

30d

Brief to formula

Not 6-18 months

30d

Brief to formula

Not 6-18 months

3

Layers of biology

In-silico, in-vitro & EEG

3

Layers of biology

In-silico, in-vitro & EEG

70%

Lower development costs

Against typical outsourced-brief spend

70%

Lower development costs

Against typical outsourced-brief spend

250 mL

Minimum order

For producers of every scale

250 mL

Minimum order

For producers of every scale

How it works

Prompt

Describe the outcome, format + the market, in plain language.

Formula

BNA™ composes FEMA-GRAS candidates + predicts how each will smell + taste.

Compliance

Market + format regulations are checked. Label copy generated with formula.

Work Order

The bill of materials opens in our ERP, priced, against live stock.

Lab

Routed to the nearest Eybna site - Van Nuys, Israel, Berlin - with 24/7 coverage.

Bottle

Ships next day, with A/B variants documented.

How it works

Three layers of biology. One closed loop.

2,200+ plant compounds modeled against 22+ human receptors - predicted, validated, and confirmed on real people.

01

In Silico

Molecular docking and dynamics predict how a compound meets a receptor.

01

In Silico

Molecular docking and dynamics predict how a compound meets a receptor.

02

In Vitro

Assays on live human cells confirm or reject the prediction.

02

In Vitro

Assays on live human cells confirm or reject the prediction.

03

EEG

Human EEG confirms the effect on real people.

03

EEG

Human EEG confirms the effect on real people.

The data

Ten layers behind every answer.

A natural-language interface. Ask it like a colleague. It answers like a formulation team.

Translucent flower rendered over a dark field, base layer of the compound map
Translucent flower rendered over a dark field, second layer of the compound map
Translucent flower rendered over a dark field, third layer of the compound map
Translucent flower rendered over a dark field, fourth layer of the compound map
Translucent flower rendered over a dark field, fifth layer of the compound map
Translucent flower rendered over a dark field, sixth layer of the compound map
Translucent flower rendered over a dark field, seventh layer of the compound map
Translucent flower rendered over a dark field, eighth layer of the compound map
Translucent flower rendered over a dark field, ninth layer of the compound map
Translucent flower rendered over a dark field, final layer of the compound map
Translucent flower rendered over a dark field, base layer of the compound map
Translucent flower rendered over a dark field, second layer of the compound map
Translucent flower rendered over a dark field, third layer of the compound map
Translucent flower rendered over a dark field, fourth layer of the compound map
Translucent flower rendered over a dark field, fifth layer of the compound map
Translucent flower rendered over a dark field, sixth layer of the compound map
Translucent flower rendered over a dark field, seventh layer of the compound map
Translucent flower rendered over a dark field, eighth layer of the compound map
Translucent flower rendered over a dark field, ninth layer of the compound map
Translucent flower rendered over a dark field, final layer of the compound map

Regulatory

Regulatory

Maps each formula to the requirements of its target territory and application.

Taste

Taste

Predicts how each compound contributes to the formula’s overall flavor profile.

Smell

Smell

Models the aromatic character, balance, and intensity of the finished formulation.

Receptor Binding

Receptor Binding

Maps how compounds interact with
human receptor systems.

Brain / EEG

Brain / EEG

Predicts the neurological response associated with the formula’s intended functional outcome.

Sensation

Sensation

Predicts cooling, warming, and other trigeminal sensations produced by the formula to build aroma.

Intensity

Intensity

Measures the expected strength of the sensory and functional experience.

Boiling Point

Boiling Point

Determines how compounds will behave under heat in production and use.

Solubility

Solubility

Determines how effectively the formula disperses across different product formats.

Color

Color

Anticipates the formula’s visual appearance and impact on the finished product.

WHAT’S NEXT

BNA™ formulates and validates.
Model Torii will predict.

BNA™ is the operating layer: brief to bottle, with receptor and EEG validation in the loop. Beneath it we are building Model Torii, a multimodal model of molecule-to-human-state biology, so that one day the prediction of how a formulation will make a person feel comes before anything is mixed. Under development, and already shaping how we formulate.

Woman with eyes closed, glowing aroma lines across her face

Let's engineer yours.

Tell us the feeling you want your product to deliver. We'll come back with a BNA-engineered formulation.