Skip to content
Julien Fleuridas
FR — Passer en français
← All projects
FRUZEE

Product · E-commerce · Research

FRUZEE

Three product ranges, from suppliers to manufacturers, and everything the industry taught me along the way.

In developmentJuly 2025 — ongoingFRUZEE

In short

Problem
Launching food products means handling research, suppliers, recipes, packaging, pricing and distribution all at once. With a tiny team, and never with all the information.
Solution
Moving range by range, from the idea to the manufacturer's spec, with a wide supplier network and AI to move faster on research and visuals.
What it brings
Three ranges developed, two live brand sites, one range ready for manufacturing and 5,000 units in pre-sale.
Where it stands
Ongoing. Fruzee and Bolis have live sites, Bolis is not on sale yet, and Zium is ready for manufacturing.
My role
I handle the offer, suppliers, costs, formats and positioning, and the commercial relationship with more than 70 suppliers and partners.

The project

FRUZEE is the company I co-founded in 2025, built around three ranges:

  • Fruzee, frozen fruit in pots that stays soft;
  • Bolis, frozen fruit bars;
  • Zium, energy and electrolyte gummies.

Fruzee was also my thesis project at emlyon. It made it to the final eight of the school's Coup de pouce competition, in October 2025.

My role

I am a co-founder, on the product and commercial side. In practice:

  • I approached more than 70 suppliers and partners, compared their prices, capacity and lead times, then negotiated launch terms;
  • I handle the commercial relationship end to end, from first contact to follow-up;
  • I choose which products to push based on the market, feedback and what each unit sold brings in;
  • I turn those choices into concrete requirements for manufacturers.

The real challenge: industry

For Fruzee, I looked for fruit suppliers, then for companies able to put the fruit in pots. The product held up, the calendar much less so: fruit and potting lines are not available at the same time. Seasonal fruit does not wait for a line to free up.

No market study shows that kind of problem. You discover it by talking to suppliers, and it often decides what is actually feasible.

What AI changed

It saved me a lot of time upstream: understanding a category, comparing brands, preparing a spec or a supplier comparison. Much less on packaging visuals, which often had to be redone. And it never decided for us: positioning, pricing and supplier terms remain human choices.

Where it stands

Two live brand sites, 5,000 units in pre-sale, one range ready for manufacturing and a negotiated supplier network. Bolis and Zium are not on sale yet, so there is no revenue to show yet.

Numbers

3

ranges developed

Fruzee, Bolis and Zium, each at its own stage

Source · public sites fruzee.fr and bolisbar.com, Zium store theme repository

What I did

  • Research

    I studied needs and competitors across three very different product categories.

  • Needs definition

    I wrote the specs for suppliers and manufacturers: capacity, cost, minimum quantities, lead times, formats.

  • Product design

    I worked on the positioning, the range, the formats and the packaging direction.

  • Steering

    I approached suppliers, negotiated launch terms and made trade-offs between the three ranges based on the market, feedback and what each product brings in.

  • Assisted development

    I created and improved the brand websites, with help from AI.

Evidence

  • Two live brand sites

    verified

    Fruzee and Bolis each have their own site, with their identity, range and story.

    Source
    public sites fruzee.fr and bolisbar.com
    Verify with
    Both sites checked on 30 September 2026: they respond.
    Two live brand sites

  • The Zium store, in a versioned repository

    verified

    Zium's online store is a versioned Shopify theme, with a stable version and a separate modernisation version. It is not live yet.

    Source
    theme repository, FRUZEE GitHub organisation
    Verify with
    Read the repository and its branches on 30 September 2026.

Limitations

  • The commercial numbers (suppliers approached, pre-sales) come from our internal tracking: nobody can check them from the outside.
  • Bolis and Zium are not on sale yet. No delivered volumes or revenue to show for now.
  • AI made no decisions for us. It saved us time upstream, much less on visuals.
  • It is a team project: the results are not one person's doing.

What I take away

  • A product can be good and still impossible to make at the right time. You discover that kind of constraint by talking to suppliers, not in a spreadsheet.
  • AI saves a huge amount of time understanding a market, comparing competitors or preparing a supplier call. Much less on visuals.
  • In an entrepreneurial project, the hardest part is never producing: it is deciding fast without all the information.

What it could bring to a company

  • A product team can reuse the same approach (market, comparison, positioning, spec, mock-up) to test a launch before committing resources.
  • A purchasing team can compare its suppliers the same way, on capacity, cost, minimum quantities and lead times.

Technologies

  • WordPress
  • Shopify
  • Image generation and editing tools

Links

Related projects