Case Study - Chatbot for Dialogue with Technical Manuals

In this project, we acted as an external party and developed an AI-driven chatbot as a Proof of Concept, specifically designed to facilitate service technicians' access to extensive technical manuals. By using Retrieval-Augmented Generation (RAG) technology, we succeeded in optimizing the search process in manuals that spanned thousands of pages.

Year
Service
Machine Learning & AI

The Project

As a partner to a provider of technical information, we took on the challenge of creating an AI chatbot for a major player in the machinery manufacturing industry. Our goal was to develop a solution that enables quick and efficient information search in comprehensive technical manuals, where the target group was service technicians.

The scope of the project was limited to a couple of service manuals. This laid the groundwork for a smooth and focused project, where we could concentrate on quality and user-friendliness and prove the solution's usability on a, to begin with, small scale.

Our solution was based on RAG technology (Retrieval-Augmented Generation), which combines a search engine with an AI chatbot. By breaking down the manual's content into smaller segments and indexing these in a vector database, we were able to create an effective method for finding and presenting relevant information.

Our work resulted in a functional chatbot that could handle specific requests from an extensive manual. Despite the project's limited scope, the results demonstrated the potential of using AI to improve information access and efficiency for service technicians.

Learnings

The greatest insight from this project was the importance of high-quality data. The success of an AI-driven solution heavily relies on the quality of the underlying data. Close collaboration with the client is crucial to ensure access to and processing of relevant data.

This project proves our ability to deliver innovative and practical AI solutions. Our customized chatbot technology has the potential to revolutionize information management and support technical processes in large-scale industrial environments.

Getting the chance to work with such new 'cutting-edge' technology was fantastic. We are very pleased with the outcome and look forward to continuing to develop the solution.

Jonathan Köre
Head of Machine Learning

Tags

  • Machine Learning
  • AI
  • Large Language Models
  • Data Processing
  • Chatbot
  • Vector Database

Fler kundcase

Named Entity Recognition for Medical Texts

In one project, our employees managed to improve the performance of Named Entity Recognition for Swedish medical texts.

Fortsätt läsa

A recipient portal for invoices

We helped Keeros build an invoice recipient portal that makes it easy for recipients to receive and manage their invoices.

Fortsätt läsa

Vill du se dina idéer bli verklighet?

Med Fiive som din partner får du inte bara expertis, utan ett engagerat stöd för att växa och leda din bransch.

Kontor


  • Kungsgatan 4
    411 19 Göteborg