Case study · Liderauto
Custom-built,
with AI at work.
An auto accessories store in Curitiba with 4,767 products. We built everything to measure: the website, the admin panel and an AI layer that automates operations and customer service.
- Client
- Liderauto Acessórios
- Sector
- Auto accessories
- Catalog
- 4,767 products
- Delivered
- Website, admin panel and AI layer
Testimonial
In the client’s own words.
Jeremias, owner of Liderauto, talks about the work in front of the store in Curitiba (in Portuguese).
In the video he mentions Rogério Costa, Kadelta’s founder. That is how we work: the client talks directly to the person who designs and builds.
01The starting point
Lots of manual work, little visibility.
- 01A WordPress site with a large catalog that was hard to maintain.
- 02Product content and descriptions written by hand, one at a time.
- 03Customers arrived on WhatsApp with no history: where they came from and what they wanted.
- 04Service, quotes and sales in different places, with no single view.
02Before and after
Same address, a different store.
Drag to compare. On one side, the old WordPress site; on the other, the new one: fast on mobile, with search by what the customer needs and the AI advisor.
BeforeAfter03What we built
Everything custom-built, with AI where it matters.
01 · Site
A new site on Nuxt
All 4,767 products migrated with URLs identical to the old ones, so nothing Google already knew was lost. Fast on mobile and installable as an app.
02 · Panel
An AI-powered admin panel
AI drafts product content and SEO; a person reviews it before publishing. The store edits everything without depending on anyone.
03 · Service
An AI advisor on the site
Understands what the customer describes, finds the right product in the catalog and hands the conversation to WhatsApp with the context.
04 · Operations
AI organizing conversations
Every conversation is read and classified: purchase, parts, after-sales, supplier. The team sees what matters first.
04What is in the dashboard
Products, conversations and results in one place.
Marketing brings the customer in; the dashboard receives, identifies and follows them to the sale. The store and the media partner look at the same number.
01
The customer arrives identified
Visitors from an ad reach the site with their source recorded, and the WhatsApp conversation carries a service code.
02
AI reads every conversation
It classifies the reason and separates purchases from after-sales or supplier messages.
03
The team follows the funnel
New conversation, quote, visit and purchase, stage by stage.
04
Results show up by source
With each campaign’s cost connected, the dashboard shows cost per sale.
Visits
Conversations
Sales
Cost per sale
Conversations
05What has already changed
Measured, with dates.
81 of 81
visits from ads reached the system with campaign and keyword identified
24 h to 04 Oct 2026
6
ad campaigns followed side by side on the dashboard
04 Oct 2026
4,766
URLs kept identical to the old site
checked 26 Aug 2026
5,038
pages published in the sitemap on launch day
29 Aug 2026
The dashboard is still gathering data. Funnel stages depend on the team moving each conversation in the service desk, and cost per campaign comes in with the Google Ads connection. Sales by source will be added here once 30 days are measured.
Does your operation still depend on manual work?
Start by measuring where you are today.