Speeding Up Maintenance Process with RAG Chatbot for Equipment Rental Company

Company Name

Location

Finland

Services

Client

Renta Group is a full-service machine and equipment rental services company that offers innovative solutions to make the rental process easier for customers.

Challenge

As part of the machinery maintenance process, Renta’s mechanics often search for particular information in a specific technical manual. With a wide range of machine types and models spanning different release years, each with distinct technical instructions and components, finding the necessary information can be, time-consuming and challenging.

Solution

DataArt’s team developed a RAG chatbot based on AWS Bedrock foundation models that streamlines and simplifies the current search process. The mechanics can ask the chatbot questions about maintenance procedures, receive answers in the form of solution summaries with links to specific pages in PDF manuals, and review the provided documents.

The primary challenges were determining the right approach for indexing documents and navigating the specific domain of the provided documentation. Also, due to regional GDPR constraints, only one LLM (Claude) was available, limiting the ability to test multiple models. With the help of a hybrid search approach (full-text and vector search), experimenting with indexing techniques, and prompt engineering, the team managed to achieve successful results and overcome technical difficulties.

Technologies

Bedrock

OpenSearch

ECS Fargate

MySQL

Lambda

DynamoDB

Outcomes

50% faster PDF searches with the chatbot compared to manual methods.
78% relevant chatbot recommendations
Increased productivity across the company

Please provide your contact details, and we will get back to you promptly.

First name*

Last name*

Email*

Job title

Company*

Phone

What would you like to discuss?*

How did you learn about DataArt?

What is your role?

I agree to receive DataArt's monthly marketing newsletter.

I have read and agree to the Privacy Policy*