BrazilRetail and e-commerce2025 – ongoingGimba
AI-assisted product catalog registration
Gimba cut product registration time from 13 minutes to 2 minutes with generative AI
Problem
Staff manually standardized manufacturers' product data and wrote catalog descriptions. Earlier experiments with AI still required manual prompts and revisions.
Approach
Gimba worked with Flexa Cloud and AWS to build a catalog tool. It used Claude 2 through Amazon Bedrock, with prompts adapted to Gimba's writing style. The registration team used one interface to produce descriptions with HTML tags.
Results
Takeaway
Standardized prompts and formatting shortened the registration team's work on new products. Reworking the existing catalog remained a next step.
Categories
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
AWS quoted Gimba's registration manager reporting that registration time fell from 13 minutes to 2 minutes per product. The page carries no date; the Internet Archive first saved it on September 18, 2025.
Full details are in the original sources.
