Miyu – the central language intelligence platform with integration capabilities
Supporting the customer service center through AI-powered automation processes.
Project profile
- GoalAI-based automation of the customer service email inbox with backend processing of enquiries
- Duration3 months
- SystemIBM Cloud, Watson Knowledge Studio and Machine Learning, further Watson services
- IndustryFashion
- TypeB2C
- Onboarding2017
Our services
Reducing Customer Service Response Times – Increasing Customer Satisfaction
An international fashion company with locations in 40 countries and a strong focus on customer-centricity faced the challenge of optimizing its digital customer communications. In numbers, this means that 600,000 emails per year tie up customer service center resources. Valuable resources are being tied up handling mostly routine inquiries—resources that would actually be better utilized for value-added tasks.
The Challenge
An international fashion company with locations in 40 countries and a strong customer-centric focus faced the challenge of optimising digital customer communication. In figures: 600,000 emails a year tie up the resources of the customer service centre. Important resources that could be put to better use on value-adding tasks are tied up by what are mostly standard enquiries.
The solution
A system was built to support the customer service centre, with AI-based analysis and classification of incoming emails, automated responses to customer enquiries based on predefined text modules, and integration into the company's backend systems to carry out the actions needed to answer and fulfil the customer's request.
Via an integration layer, Miyu accesses the relevant backend systems and thus speeds up laborious processes such as gathering information, generating documents or updating records.
Miyu uses feedback features to continuously improve the machine learning algorithms it employs, which are seamlessly integrated into customer service processes and can be used by customer service agents.
The Technology:
Miyu uses feedback functionality to continuously improve the machine learning algorithms it relies on; this functionality is seamlessly integrated into the customer service processes and can be used by the CS agents.
The technology
Miyu was implemented in the IBM Cloud. Feature extraction and training were carried out using IBM's Watson Knowledge Studio and the text mining of its natural language processing.
A neural network in Watson Machine Learning was used to classify the emails.
The application was built with Node.js in the IBM Cloud.
Let's talk about your project.
Every project you see here started with a simple conversation. Send us a quick note explaining what it's about. We'll figure out the rest together.
Please accept the corresponding cookies to view this embedded content.