top of page
Header BG Case-study.png

Japanese truck manufacturer gets faster issue resolutions using NLP

Google Cloud Data Analytics.png
Google Cloud ML specialisation.png
Google Cloud Premier Partner.png
95% NLP accuracy in identifying top historical issues

Client

The client is a leading Japanese original equipment manufacturer (OEM) of commercial trucks, serving both domestic and international markets. With a large service network and complex quality assurance processes, the company needed to accelerate issue resolution by leveraging AI and multilingual NLP capabilities.

Project Context

The client wanted to speed-up the resolution of quality issues by using historical quality and solutions. New issues might be reported by dealers but searching and matching them was hard especially when issues are reported in both English and Japanese.

Challenges

Translated Japanese issues into English for consistent analysis.

Used a BERT-based NLP model to match new issues with historical ones.

Suggested top three relevant past issues for faster resolution.

Replaced manual search with automated, accurate recommendations.

Solved both language challenges and time delays in issue handling.

Solution

Dealers reported fresh quality issues frequently, but matching them to historical problems was time-consuming.

Quality issues were logged in both English and Japanese, making search and comparison difficult.

Quality engineers had to manually sift through large databases, delaying resolutions and reducing customer satisfaction

Project Objectives

Decrease the time spent by Quality engineers in searching issues.
Increase customer satisfaction and better manage quality.
Resolve issues in both English and Japanese.

Solution Delivery

Japanese issues were first translated into English for uniform processing.

A custom BERT algorithm compared new issues with historical data.

Top three matching issues were surfaced for engineers to reuse or adapt solutions.

The NLP model was integrated with a custom UI and backend on Microsoft Azure.
Quality issues were logged in both English and Japanese, making search and comparison difficult.

Quality engineers had to manually sift through large databases, delaying resolutions and reducing customer satisfaction

Testimonial

Thanks to SquareShift’s NLP engine, we can now resolve quality issues faster and more accurately across languages

Technology Stack

To know more in detail 

bottom of page