A Technical Advisor, an S/4HC Search Engine, a Meeting Coordinator, and an Entertainer!

Edmin Autobiography – the Complete Guide

Edmin-personalities.png 

According to the Myers Briggs theory, there are sixteen different personality types – and these are prevalent among chatbots too! For those who do not already know me, I am Edmin, the S/4HANA Cloud Community Mascot. I have many strengths and – not so many – weaknesses. This is because, not only can I address S/4HANA Cloud and community related questions, but I can also retrieve SAP articles, raise Customer Meeting Support Requests, and I might also make you laugh. I am not your average digital assistant… I am a Technical Advisor, an S/4HC Search Engine, a Meeting Coordinator, and an Entertainer!

So let’s explore some of the options that will help you make most of your experience with me.

 

1 1)      The Main Chat Operations

chat1.pngchat2.pngchat3.png

One way of initiating the conversation is through greetings. You will then be provided with the three main operations you can execute:

  •         Ask a Question: Ask me a question and I will see if I have a relevant answer for you. If you are happy with the answer, great! If not, then I will execute a search.

 

  •           Search: Are you looking for information around a specific topic? I can do a quick search for you and provide you with a list of SAP resources to review. Please provide me with your search text only. If you fail to do that, I can trim your sentence down to the technical keywords to optimise your searching experience. You can always decide whether you want me to retrieve results based on your original text or my suggestion, simply by pressing ‘Fulltext’ or ‘Keywords only’.

chat4.pngchat5.png

  •           Help request: Do you need live support? You may now raise a customer support request. You will be asked to provide a subject, name, and e-mail and an expert will reach out to you to arrange a meeting upon your request. As this initiative is in pilot mode, please note that we currently accept requests in Procurement, and Sales and Distribution, and can also support you with topics around Integration and Data Migration.

For further information on the main chat operations, click here.

2 2)      Quick Menu with Shortcuts

chat6.pngchat7.png

A menu icon with shortcuts is also available as an alternative to writing replies. ‘Ask a Question’ and ‘Search’ will post a message to the bot that will trigger the equivalent process if pressed. ‘Help Request’ branches off to ‘Create Meeting Request’ and ‘Cancel Meeting Request’, which will also trigger the relevant processes if pressed, while the bottom two options are hyperlinks to Community blogs if you require assistance.

3 3)      Small Talk

chat8.pngchat9.png

You may skip the greetings and the small talk and start off with your question or search request, but for those who like to chat, I am up for it too. Always happy to catch-up with you. However, if you try to fool me, you will sooner than later realise that I am not having any of it…!

4 4)      Feedback

chat10.pngchat11.png

Near the end of your conversation, depending on whether you were happy with the provided solution or not, you will be either encouraged to post a question to one of the community forums or you will be asked to provide some feedback. Feedback is important for me as it influences my development as a chatbot.

5 5)      Behind the Scenes…

Finally, I would like to give you an insight on what is happening behind the scenes. The solution has been developed on the Conversational AI Platform that uses Natural Language Processing (NLP). The NLP technology is a branch of artificial intelligence that examines the way computers interpret human language.

But where does the artificial intelligence take place? While humans use intuition and context to make sense of language, computers have difficulty in coping with the ambiguity of it. This is because meaning does not only depend on words alone, but on the combination and sequence these follow. For example, ‘I want to get an apple’ and ‘I want to go to the Apple store’ demonstrate the importance of context in interpreting language. For this same reason, NLP technology analyses syntax and semantics.

In simple terms, chatbots need to be told what sentences to expect and what keywords they need to look out for in order to trigger a certain response. They also need to know different ways of expressing the same meaning. For every phrase you type, the chatbot assigns your phrase to a category of similar expressions, and the more synonym expressions you provide to the bot, the better it copes with language ambiguity. To cut the long story short (although I would happily go on and on…), the better the training of the chatbot, the better the performance of NLP.

If you are curious to know what the conversational flow looks like, here is a simplified version of it:

Edmin-flow.png

6 6)      Product Roadmap

For the next release, we are planning to improve the search functionality by integrating to the API Hub and SAP launchpad.

 

Would you like to influence the product Roadmap? We are always happy to hear your feedback! Please send an e-mail to eva.stergiou@sap.com.