What challenges did Grupo Elfa face before AI implementation?

Manually processing high volumes of quote requests by email was slow and inefficient. Salespeople were spending significant time on data entry instead of customer relationships and selling. The company needed to interpret unstructured customer communications and apply complex pricing, tax, and logistics rules.

What was the impact of the AI solution?

Quote processing time was reduced from 40 minutes to about 2-3 minutes per 100-item quote. The solution generated more than 100 million Brazilian reais in new revenue within 8 months. The system achieved 99.2% accuracy. Salespeople could spend more time on customer relationships and negotiations.

Rank#190of 3,107

Source

Grupo Elfa used Amazon Bedrock and Amazon Textract to automate healthcare quote processing

Use case typeDocument automationUpdated Jun 13, 2026

Grupo Elfa, a Brazilian healthcare distributor, worked with A3Data and AWS to automate the processing of thousands of daily customer quote requests. The company built CotAI to reduce manual email-based quote handling and help sales teams respond faster while improving accuracy and customer experience.

Organization: Grupo Elfa
Industry: Healthcare
Location: Brazil
Published: April 2026

Reported outcomes

99.2%

accuracyQuality & accuracy

40 minutestime3 minutestime

Strategic outcomes

Speed & agilityAutomated quote processing workflowNew product / capabilityInterprets unstructured customer communicationsCustomer experience & trustImproved customer relationships and negotiationsNew business modelGenerated new revenue from automation

Catalog median for quality & accuracy deployments: +90% across 282 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1Document automation
2Sales operations automation
3Quote generation

Manually processing high volumes of quote requests by email was slow and inefficient.
Salespeople were spending significant time on data entry instead of customer relationships and selling.
The company needed to interpret unstructured customer communications and apply complex pricing, tax, and logistics rules.

Grupo Elfa and A3Data developed CotAI using Amazon Bedrock for foundation model access.
The solution uses Amazon Textract to extract text, handwriting, layout, and data from scanned documents and unstructured inputs.
AWS Lambda supports serverless processing and Amazon S3 is used as part of the data lake.
CotAI reads incoming emails, determines whether they are quote requests or orders, extracts items to be quoted, creates CRM opportunities, and helps determine pricing, tax, and logistics options for human review.

Technologies

Amazon Bedrock Amazon Textract AWS Lambda Amazon S3

Quote processing time was reduced from 40 minutes to about 2-3 minutes per 100-item quote.
The solution generated more than 100 million Brazilian reais in new revenue within 8 months.
The system achieved 99.2% accuracy.
Salespeople could spend more time on customer relationships and negotiations.

Architecture

CotAI uses Amazon Bedrock for model access, Amazon Textract for information extraction, AWS Lambda for serverless processing, and Amazon S3 for data storage and lake-style management; the workflow ingests emails, classifies quote requests, extracts key items, and routes results for human review.

Implementation partners1

A3Data

Sources & evidence1

Groundedness: 5/5

Primary source

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Customer evidence