Chatbots for process optimization
In the 8th episode of the Chatbot Talk, Sylvie Pflitsch from Bosch is our guest and the topic is chatbots for process optimization – (bots for optimizing processes).
Sylvie Pflitsch and what she loves about chatbots
Sylvie is a process manager and leads the digital transformation at Bosch Security Systems & Building Technologies in Grasbrunn near Munich. She describes herself as a process manager with a passion for optimization. Your department is the bridge between IT and business, so to speak.
Sylvie saw the potential of chatbots for business straight away. For them, chatbots are business enablers because they can really do a lot. Chatbots are among other things
- Available 24/7
- relieve the workload
- have a high level of user-friendliness if they are designed accordingly
- implemented relatively quickly
- can trigger other databases or processes
Sylvie knows how important design is for a user-friendly chatbot and that it is helpful to have a concept when creating a chatbot. The concept helps with topics such as what might interest the user, how the communication flow works, how we can design the buttons, etc. step by step. This makes the chatbot user-friendly and fits the target group.
Bots at Bosch – chatbots for process optimization
At Bosch Security Systems, Sylvie and her team mainly create chatbot “prototypes” to test and demonstrate the effectiveness of chatbots. Sylvie describes the chatbots as prototypes because chatbots are never really finished and can always be optimized.
Chatbots can be used in customer communication or for internal communication. Design workshops and prototypes are used to familiarize employees from the specialist departments with the topic of chatbots and take away their fear of chatbots. Together with the specialist departments, we are looking at where chatbots can be used sensibly. At the moment, the chatbots are being used internally in the Q&A area to speed up processes.
The usable chatbots are created by the specialist departments themselves using various tools.
What is the process of a chatbot project?
Sylvie and her team sit together with the respective specialist departments and together they go through the individual processes. Questions like “What do you want to do? In what context? What is your idea?” are discussed and the thoughts are first put on paper.
Then we move on to the use of chatbot tools. The employees are shown how the chatbot tools work and then the first flows (chatbot dialogs) are created. When creating the flow, the process questions come up again. Sylvie and her team have found that errors in the concept are discovered more quickly if you start designing the dialog early on and don’t get lost in too many details at the beginning. “You can see straight away that things you imagined would work well don’t.”
And what about the chatbot personality?
The chatbots at Bosch have a name, greet and say goodbye to the user, but otherwise have no great personality. So they are rather sober. Until now, chatbots have tended to be fact-oriented. You know the tone of internal communication and just as you would reply in an email, you can also let the chatbot reply.
When asked if they would do it differently if they were to make the chatbot outwardly for customers, Sylvie replies that she would definitely do it differently. Then you have to look at which customer group, how this group communicates, etc.
The two biggest learnings from Sylvie and her team
- It is often not a tool problem. If it hangs, then it’s usually not the tool, but the process. It’s always about the process and the content. It’s then about how I can adapt the process so that it works.
- A chatbot brings no added value if nobody knows it. A chatbot must be made known.
What does the future of chatbots look like at Bosch?
The chatbots at Bosch Security Systems are currently rule-based. They are still gaining experience with the use and acceptance of chatbots. Of course, technology offers much more, but in their field they still have to learn how to deal with high-performance technology.
The aim is to develop bots whose knowledge base becomes more intelligent. They deal with neural networks and how they can develop the answers themselves on the basis of existing question & answer lists.
It’s best to listen to the episode yourself and get to know Sylvie Pflischt personally. Have fun![vc_empty_space height=”40px”]
[vc_empty_space height=”40px”]And if you don’t want to miss any more podcast episodes, then sign up for the mailing list at the bottom of this page.
