robots walking in San Francisco
Three chatbot demonstrations brought the theory of conversational interfaces to life. PHOTO: Jere Keys

Chatbots have invaded the workplace. Or at least, invaded the conversations around the digital workplace.

In the last month, I have seen some outstanding presentations on the use of conversational interfaces (aka chatbots, aka bots) in the digital workplace/intranet context. While a number of interesting articles have been published on the topic in the last year, the presentations had a hands-on, “lets write a simple bot” element to them that made the abstract real. 

What follows is an overview of some of these presentations to get you thinking about how bots might benefit your digital workplace. We'll start with a complex "artificial intelligence" (AI) driven platform and end with a simpler introduction to the benefits and the potential value.

Simplifying Chatbot Creation

Chris McGrath walked attendees of the JBoye Digital Workplace group meeting in Toronto through the operational concepts of conversational user interfaces in internal communications as well as the broader digital workplace. When I first met McGrath over 10 years ago, he was co-founder of Vancouver, British Columbia-based ThoughtFarmer, provider of a social intranet product. McGrath is now founder of chatbot company Tangowork

McGrath's non-salesy demo featured a real bot developed for the Intranet Now conference that used Facebook Messenger as the channel. He also showed examples of other real-life applications for bots that used either Skype for Business, SharePoint web parts or Microsoft Teams as their interface layer. The demo was fascinating, as it showed coding these as interfaces was easier than I thought. 

Here's an overview of the high-level benefits of the kind of chatbot McGrath demonstrated:

  • No Installation — You already have the messaging app.
  • Instant Upgrades — Server-based apps.
  • Fast — Text does not eat up bandwidth.
  • Read in Minutes — Push notifications.
  • Easier to Read — The medium dictates brevity.
  • Easier to Write — That brevity thing again.
  • Accessible — Text and voice user interfaces can benefit users with disabilities.
  • Works Without Internet Connection — Can use SMS for text.
  • No Need for User to Context-Switch — Access information in your message app.
  • Ease of Use — If you can type or talk, you can use a chatbot.

McGrath also discussed how natural language processing (NLP), as part of the AI tool set, enables the back end processing of queries. NLP provides a segue into another great chatbot session I saw.

Chatbots Rely on Clean Content 

UK-based consultant Sharon O’Dea ran a session called, “Taking conversational design to the next level” at the JBoye conference in Aarhus, Denmark in early November. 

O'Dea shared insights from her work for both commercial and government clients, and showed examples of chatbots used for meeting room booking, expenses approval, vacation booking, delivered through a variety of interfaces including Microsoft Teams, an intranet page and of course, a mobile messaging app.

Both O’Dea and McGrath emphasized the need to clean up your content before embarking on a chatbot program. If you want a chatbot's natural language processing back end to provide good answers to your employees, you'll need to put some effort into content curation up front. The old adage, garbage in — garbage out, once again holds true here. As a test, enter your content from an Intranet FAQ by ingesting it into Microsoft’s free bot framework and see the results. 

Microsoft Bot Builder SDK

Natural Language Processing in Action

The final chatbot demo came from the Deloitte team in Toronto, who hosted a session on future technologies for Legal and Procurement functions at their Greenhouse innovation lab. One of the technologies they demonstrated was the Amelia AI system from New York City-based IPSoft. Amelia is more than a chatbot system, with other cognitive computing capabilities, but it also uses conversational interfaces to receive and respond to questions. 

As part of the demo, a member of the team cut and pasted the content from Deloitte's Wikipedia page into the Amelia system. The natural language processing engine created a concept map while we watched (granted, it wasn't a huge amount of content for a demo) and within minutes we were asking questions like, “how many people work for Deloitte” and getting comprehensive answers in conversational English in the chatbot interface. A simple but impressive demo. According to the IPSoft website, Amelia can be deployed in other, more sophisticated use cases. 

Related Reading: Workplace Chatbots – Too Little, Too Soon?

Do Conversational Interfaces Belong in Your Digital Workplace?

Do you think your workplace would benefit from conversational interfaces? I am looking forward to experimenting with this technology. My organization uses Skype for Business and I can see some real benefit in lawyers typing in a question to a “Knowledge Management Chatbot” rather than searching our intranet portal for the information they need (even though it is well organized, if I don't say so myself). 

The next level is voice access to a chatbot, through Alexa, Siri, Cortana or Google Assistant. One way or another, we will be conversing with our digital workplace in the not so far off future.

If your organization is already speeding ahead and using some form of conversation interfaces, please leave a note in the comments section and let us know what you're doing.