Artificial intelligence is pushing the boundaries of personalization and redefining the relationship between consumers and their fragrances. The goal: to allow everyone to create a perfume that reflects their personality, tastes, emotions, and identity.
Ininum: Customize your perfume in a few taps
Slovenian start-up Ininum is bringing fragrance personalization to consumers through a mobile app connected to a compact device containing three perfume-base capsules. The concept is simple: create a bespoke fragrance for any occasion. Users can either describe their needs through the company’s AI-powered chat interface or upload a photo of their outfit. After analysing the request, the AI recommends a tailored fragrance blend designed to match the context, mood, or style.
Targeting the B2C market, Ininum is now launching its second product range following strong traction in the United States. The complete system retails in Europe for EUR 299. Developed in partnership with fragrance house Moelhausen, the three accords at the heart of the device took five years to perfect and can be combined in countless ways. The app also incorporates a social dimension, allowing users to share their creations, and exchange feedback with the community.
Signatures Parfumes: Custom layering
Powered by an award-winning AI app recognized at the Digital Beauty Awards, Signatures Parfums offers a patented technology that enables consumers to create their own bespoke fragrances using five fragrance-base capsules. The system can generate more than 75 million possible combinations. The device relies on highly precise layering technology, dispensing each accord in exact proportions down to the millimetre without mixing the capsule contents. The fragrances only merge once sprayed simultaneously on the skin.
Signatures Parfums is one of the brands developed by startu-p Scent-Technologies. The system was conceived by Armando Medina, founder of AirParfum, the scent-delivery technology company later acquired by Puig.
Deepscent Inc., Fragrance as a Service (FaaS)
South Korean start-up Deepscent has developed an AI-powered olfactory content creation platform capable of translating environmental data — such as images and music — along with a desired mood or atmosphere into scent. The technology enables users to generate personalized fragrances on demand within seconds. Built around the concept of Fragrance on Demand and Fragrance as a Service, the platform connects to a scent diffusion module called the Deepscent Lounge. Equipped with four fragrance capsules, the system can create up to 256 home scent combinations. The company currently offers a catalogue of around a dozen capsules.
Recognized with a CES 2026 Innovation Award, Deepscent is now targeting the European hospitality, museum, and smart building sectors. Its solution is designed to integrate directly into building management and monitoring systems.
i-Oasis: Expanding possibilities
Immersive training solutions based on virtual reality (VR) and extended reality (XR) are at the heart of i-Oasis’ business model. To enhance realism, the company has developed a catalogue of scents that can be synchronized with virtual experiences. Its lightweight Aromadive device attaches directly to a VR headset and can diffuse up to 10 fragrances during training sessions lasting only a few minutes.
Founder Abdoul Benamer is now expanding the technology’s applications beyond training and into ambient scenting.
“The connected Aromadive model we are showcasing at VivaTech can be installed in a hotel room and programmed to diffuse different scents throughout the day: relaxing notes when guests arrive from the airport, the aroma of freshly baked croissants in the morning, or a more intimate atmosphere in the evening,” he told Premium Beauty News.
By leveraging guest data — such as country of origin, travel profile, or whether children are accompanying them — the system can tailor olfactory experiences to individual preferences. The technology can also adapt scents in real time according to room activity levels detected by integrated sensors. The concept is particularly aimed at luxury hospitality, where hotels could offer highly personalized scent experiences while building a memory of guests’ preferences for future stays.
“A simple QR code allows guests to download the intuitive app and customize the scent schedule themselves,” adds Abdoul Benamer.
EveryHuman: Introducing more subjectivity
A pioneer in algorithmic perfumery, EveryHuman has expanded its offering beyond its original online questionnaire by introducing a conversational AI chat interface. Users can now describe memories, or more detailed personal stories in a natural dialogue format. This qualitative input is then processed by the algorithm, which translates the data into a fragrance profile that is subsequently executed by a robotic production system.
Lylo AI by Le Studio des Parfums
Le Studio des Parfums Paris, co-founded by Patrice Dana, has developed Lylo AI, a formulation support tool built on a proprietary fragrance database. Lylo helps simplify the understanding of more than 1,200 raw materials used in perfumery. Based on a short questionnaire covering lifestyle and olfactory preferences, the system generates two tailored fragrance recommendations.
Le Studio des Parfums also collaborated with Amazon on The Fragrance Lab project in 2025.
Amazon’s Fragrance Lab
A standout feature at the Amazon stand was The Fragrance Lab, where the promise “Design my fragrance” was brought to life using Amazon Bedrock, the company’s platform for building generative AI applications and agents.
The experience begins with a series of questions displayed on screen. Equipped with a headset and microphone, users engage in a natural voice conversation powered by Nova 2 Sonic AI, Amazon’s proprietary speech model. The system then translates the collected input into a coherent fragrance formulation.
The final step is to generate a visual communication campaign including a visual and a base line. At VivaTech, the assembly of perfumes from the bases was carried out the old-fashioned way, with pipettes. Will the American giant go further in olfactory intelligence by automating the process?
