Renault called on Dydu to deploy an IT chatbot for its employees. Ali Skalli, ITSM and Self-service Product Manager, looks back at the implementation of this project and shares the next steps for the future.
Overview ✅
🎯 GOAL: to reduce helpdesk requests by setting up an IT chatbot accessible via Teams.
Since going live in mid-2021, the bot has recorded:
- +72% increase in traffic thanks to the integration of the bot into Teams in October
- 137,000 visitors
- 36 K interactions
- 94% comprehension rate
Can you describe Renault and your role?
Renault is an innovative company that offers services for customers (B2C), as well as high-performance internal solutions (B2B).
I am Ali Skalli, ITSM and Self-service Product Manager in the DIR-ICS department (Renault IT Department – IT Cloud & Services).
What is the role of Selfcare tools, particularly chatbots, at Renault?
Previously, employees had to call the helpdesk to report incidents. This model was efficient, but it also had its drawbacks.
At Renault, we set up an IT user portal to:
- Order a product or service
- Solve problems by yourself, in just a few clicks
- Report an incident via a specific form
- Since 2018, a chatbot is also available to facilitate user interaction and help employees solve their problems.
Which issues did you want to address by implementing a chatbot?
Employees would call the helpdesk to ask questions, even though most of the answers could be found in our existing documentation.
BotIT (the name of our chatbot) was designed to provide direct answers to IT-related questions and make it easier to access our documentation, without needing to pick up the phone, which often led to dissatisfaction.
We then improved the bot to help users formulate their needs more precisely (using decision trees) and to provide the most suitable response.
What scope and use cases does your chatbot cover? Did you encounter any difficulties when building your knowledge base?
BotIT answers employees’ IT questions. At first, we limited the knowledge scope to desktop upgrades, detecting the most frequently asked questions and answering them automatically.
We then analyzed the questions and answers (user intents) and guided users to help them specify their requests. We also extended the bot’s scope to other uses. BotIT now covers topics such as PC migration, email delegation, VPN issues, password changes, etc.
The Dydu solution is user-friendly, so it didn’t take long for people to get the hang of it.
How did the Dydu teams support you in setting up and monitoring the project?
A Customer Success Manager (CSM) at Dydu supported us throughout the project. He helped us define the scope of our project and build the knowledge base before going live. Our CSM is still working with us to optimize and develop our project.
How many people manage the chatbot internally and what do they do?
There are currently two people managing the bot:
-
One person manages the bot’s content and acts as a bridge between the Renault world and the bot’s technical manager.
-
Another person is in charge of analyzing the dialogs and improving technical uses (e.g., API integration).
The functionalities we use the most to improve the bot are the knowledge part, of course, but also web services, dialog analysis, and analytics.
Have you taken any specific actions to boost your bot’s traffic?
We’re working on it.
The bot is now available on Teams, which has made it more visible. We’ve also integrated the chatbot into our IT Selfcare portal. We still need to step up our internal communication strategy to encourage employees to use the bot when they have IT questions.
Do you think the chatbot is a success? Why?
To monitor our bot’s success, we track the number of dialogs per month, the satisfaction rate, and the comprehension rate vs the confusion rate. The bot has a good user satisfaction rate, so we consider it to be a success.
The bot aims to support the helpdesk, which receives a lot of calls. The bot reduces the number of calls to the helpdesk.
Why would you recommend the Dydu solution?
The Dydu solution is easy to use, and the teams are quick to implement it. It meets our technical needs since it can connect with several channels such as Teams, Slack, Web App, etc., and also use external NLUs.
Low code / no code API integration is another advantage.
What developments and/or updates are you planning to improve the chatbot and encourage its use?
The next step will be to optimize our communication strategy to promote the chatbot.
We then plan to develop increasingly automated use cases, in collaboration with the IT team.
Share this article!
