Ecommerce ops · Production

Froméo AI cheese butler by Fromages d'ici and LG2 answers 99.77% of user queries with Botpress

The problem

With over 1000 local cheeses to choose from and a new scientific cheese classification system to communicate, consumers faced an overwhelming selection experience that passive browsing and static menus could not address.

Workflow diagram · grounded in source
1
Taste questionnaire entry
trigger
“Users initially answer a concise questionnaire to identify their tastes”
2
NLU-driven conversational refinement
ai_action
“By integrating advanced AI and NLU, Froméo effectively handles interactions without requiring users to understand technical terms, making the experience more intuitive and personalized”
3
Expert knowledge base lookup
ai_action
“The chatbot's knowledge base was carefully curated from proprietary content developed in collaboration with recognized experts – such as the Centre d'expertise fromagère du Québec (CEFQ) for cheese, the Association des microbrasseries du…”
4
Personalized recommendation delivery
output
“they receive quick, personalized cheese recommendations based on scientifically determined preferences”
Reported outcome

Froméo answers 99.77% of user queries and drives users to explore more site content after chatting, replacing passive browsing with an active, guided conversational experience that promotes local cheese production.

Reported metrics
User queries successfully answered99.77%
Users exploring more site content post-chatof users explore more site content after chatting
Reported stack
BotpressGPT-4oNLU
Source
https://botpress.com/customers/fromages-d-ici-and-lg2
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Froméo answers 99.77% of user queries and drives users to explore more site content after chatting, replacing passive browsing with an active, guided conversational experience that promotes local cheese production.

What tools did this team use?

Botpress, GPT-4o, NLU.

What results were reported?

User queries successfully answered: 99.77%; Users exploring more site content post-chat: of users explore more site content after chatting (source-reported, not independently verified).

How is this ecommerce ops AI workflow structured?

Taste questionnaire entry → NLU-driven conversational refinement → Expert knowledge base lookup → Personalized recommendation delivery.

Related ecommerce ops cases