AI in Fragrance Creation: Reality or Futurism?

It smells like a revolution. The fragrance industry, worth over 50 billion dollars, and consumer market giants are actively searching for new tools.

Artificial intelligence has become their most promising assistant. But does this mean that the algorithm will soon replace the perfumer? Don’t rush to say goodbye to century-old art.

Fundamentals of Perfumery

• Note — a single scent, for example, jasmine or bergamot
• Chord — a harmonious combination of notes, creating a complex bouquet
• Raw materials — from natural rose oil to complex synthetic molecules
• Brief — technical specification, often poetic («freshness of the morning after the rain»)
• Formula — an exact recipe for creating a fragrance
• Pyramid — a description of how the fragrance unfolds over time

How AI is transforming the industry

The history of perfumery began with alchemists, but today companies like Givaudan use AI, reducing the development of new fragrances from several years to months.

Technological breakthrough

Today, AI is penetrating every stage of fragrance creation:

• Prediction of a molecule’s smell based on its structure
• Generation of new compounds
• Selection of raw materials considering the budget

Ключевые игроки

They clashed on this field.ambitious startups (Osmo, основанный бывшими инженерами Google) и traditional perfume giants (Symrise, Firmenich).

Technological challenges

«To predict the smell of a mixture by knowing only the properties of its individual components is like trying to guess the taste of a complex dish by tasting each ingredient separately.»

Working tools

Scentmate by Symrise

A smart assistant offering affordable alternatives to expensive components

Персонализация Chanel

Creating a unique combination from ready-made chords through an online quiz

AI limitations

AI handles nonlinearity poorly and cannot translate a chemical structure into an emotion. Full autonomy is not yet achievable.

Economic efficiency

Development times are reduced by 70%, and costs—especially for small independent brands.

Coty created its first fragrance using AI in 5 weeks instead of the usual 18 months.

Glossary

IFRA International Fragrance Association, setting safety standards

QSAR/QSPR Methods of Quantitative Forecasting of Substance Properties

ГХ-МС (GC-MS) — Gas Chromatography-Mass Spectrometry Gas chromatograph coupled with a mass spectrometer

Checklist for an AI Brief

• ☑ Raw material constraints: number of components, budget per milliliter
• ☑ Desired character: key notes and chords, emotion
• ☑ Technical framework: compliance with IFRA standards, sustainability requirements