Grupo Elfa revolutionizes healthcare sales using AWS and A3Data

Grupo Elfa distributes medications, hospital supplies, and medical equipment to over 7,000 hospitals, 200,000 clinics, and 700 health insurers across Brazil. With a team of over 1,000 internal salespeople, the company had difficulties managing the high volume of price quote requests received daily. Each salesperson was handling an average of 30 emails daily, with each email containing more than 100 items to be quoted. This manual process was time consuming and inefficient, limiting the company’s ability to respond promptly to customer requests and potentially losing sales opportunities.

Grupo Elfa wanted to find a way to automate and streamline this quote process, letting its salespeople focus more on customer relationships and sales rather than manual data entry. The company needed a solution that could accurately interpret various types of customer communications, extract relevant information, and generate quotes quickly and efficiently while maintaining accuracy and compliance with complex pricing and logistics requirements.

To address this challenge, Grupo Elfa worked alongside A3Data to develop CotAI. The generative AI–powered solution is supported by a host of AWS services. For connecting and testing various large language models, CotAI employs Amazon Bedrock, a fully managed service that offers a choice of high-performing foundation models. To extract information from unstructured data sources, it uses Amazon Textract, a machine learning service that automatically extracts text, handwriting, layout elements, and data from scanned documents. The solution also uses AWS Lambda, which makes it possible for users to run code without thinking about servers or clusters, for serverless development. And as part of the data lake for storing and managing data, it uses Amazon Simple Storage Service (Amazon S3), an object storage service.

CotAI uses AI to read incoming emails, interpret whether they are quote requests or orders, and identify the items to be quoted. It then automatically creates opportunities in Grupo Elfa’s customer relationship management system. For each quote, the solution also determines the best pricing, tax structure, and logistics options, which a salesperson will then review to verify accuracy. This comprehensive approach provides an automated quotation process that significantly reduces the manual workload for Grupo Elfa’s salespeople.

The implementation of CotAI has delivered significant benefits to Grupo Elfa. The quote processing time has been reduced from 40 minutes to just 2–3 minutes per 100-item quote, representing time savings of up to 95 percent. In terms of revenue growth, CotAI generated over 100 million Brazilian reais in new revenue in the first 8 months following implementation. This increase in revenue is because salespeople can focus more on developing and maintaining customer relationships.

Beyond the impressive financial results, CotAI has achieved a 99.2 percent accuracy rate. Additionally, the salesforce has been empowered to focus on higher-value activities, like relationship building and negotiations, rather than manual data entry. "As we work with the purpose of really bringing health to people, from the moment you use technology to be more efficient, we are increasingly bringing more health, at a more affordable cost, with a better experience for our customers. In short, by doing that, we increasingly impact the health of our country." Grupo Elfa plans to continue developing new AI-powered solutions using AWS and A3Data, with the goal of becoming the most efficient company in their segment in Brazil.

For AWS or Grupo Elfa: This quote has been modified from the following: ""As we work with the purpose of really bringing health to people, from the moment you use technology to be more efficient, we are increasingly bringing more health, at a more affordable cost, with a better experience for our customers. In short, by doing that, we increasingly impact the health of our country." Please let us know if you approve this change, or if you'd prefer for us to use the original translated text.