Predicity — Four years of technical partnership
Four years of technical partnership, from internal prototypes to the data and AI platforms built for major beauty companies.
- Sector
- Beauty & fragrance · Data & AI
- Engagement
- Technical partnership
- Timeline
- 2022 → today
- Expertise
- All our expertise
Context
Predicity builds data and AI technologies for the beauty and fragrance industry, around consumer perception, recommendation and the analysis of business data. Since 2022, we have worked alongside them as their technical partner.
What they need is not a team with a steady backlog to work through. Their activity follows the rhythm of their own projects: some periods call mostly for challenging an architecture, assessing whether an idea is feasible, sizing a piece of work or preparing a client presentation. Then a new project lands, and a full team has to be available within days, for several weeks or several months.
Day to day, Predicity can call on us to secure a technical decision, test an idea, estimate a project or bring a software, data or AI perspective into a client conversation. That availability is deliberately light. When a project genuinely starts, we move into delivery quickly: framing, design, frontend, backend, data, infrastructure, AI, then support and maintenance where needed.
The same team keeps the context on their products, their models, their business constraints and their technical choices. Nothing has to be explained again at the start of each new project.
Our approach
The partnership rests on a simple idea: keep senior capacity available over time, but mobilise a full delivery team only when a project warrants it.
For Predicity, that means moving without friction from a two-hour feasibility conversation to a structural project lasting months.
Same team, same context, from one project to the next
Predicity keeps continuous technical availability at a limited cost. Projects are then framed and committed to separately, once delivery capacity is genuinely needed.
What we have built together
Over four years, the work has taken very different forms.
Exploring new uses
We regularly build prototypes and proofs of concept with Predicity, meant to make their models tangible for prospects, partners and clients. Spark, their emotion-based recommendation experience, is one of them.
We also designed a variant combining Spark with our own Naaro technology, to explore a fully conversational recommendation experience.
Turning models into business products
We also work on the interfaces through which the output of Predicity models is actually used. Consumer Pulse is one example: a more analytical reading of consumer perception, designed so that business teams can act on the results directly.
Resonance is another, and goes further into the craft itself: a fully digital fragrance creation tool that starts from the marketing brief, holds it against consumer perception and the references on the market, and takes it down to formulation proposals.
These projects usually combine several dimensions: understanding the need, structuring the data, interface design, application development and the integration of the AI components.
Building platforms for large accounts
Some of this work becomes a months-long project run directly with Predicity clients. Vezero then covers the whole chain: understanding the need with the end client, framing, UX/UI, mockups, development, delivery, then support and maintenance.
Among those projects — whose brands cannot be named — is a market analysis platform for a major American beauty company: a customised version of the Predicity technologies, which explores thousands of products on the market through consumer perception, millions of reviews, olfactory notes and launch dynamics.
And an analytics platform for a French perfume house: one tool bringing together the product catalogue, the metadata produced by the creation and innovation process, consumer and expert testing, ingredients and olfactory notes, with assisted report generation to support internal and external presentations.
The platforms built for Predicity clients cannot be shown. The interfaces above are those of their own products.
What the partnership makes possible
These projects differ widely in length, context and complexity. What does not change is the team.
Four years of accumulated context is what allows a feasibility conversation to become a full project without starting from scratch: we know the products, the models, the technical choices, the business constraints and the way Predicity works with its own clients.
That continuity sharply reduces the time it takes to frame a new subject, secure a decision or start a piece of work. It is exactly what the technical partnership is meant to produce: a senior team available over time, but mobilised for delivery only when it is genuinely useful.
What Predicity said
Predicity regularly calls on Vezero to deliver our bespoke data and AI solutions. Their team gets straight to the point, frames quickly, executes cleanly, and always keeps the end use in mind. The result: reliable deliveries, impeccable quality, and genuine peace of mind on the client side, even on demanding projects.
Do your technical needs vary over time, while you want the same team on hand whenever a project opens up?