Person dressed up in a robot costume typing on a computer in a cubicle.
Robots are, in fact, taking over the digital workplace — in a good way, for many. PHOTO: Ben Husmann

David Cearley, vice president and Gartner Fellow, wrote that promises of artificial intelligence (AI) magically performing intellectual tasks that humans do and dynamically learning as much as humans is "speculative at best." However with 2018 rapidly approaching, AI is clearly on the minds of many businesses. Where are businesses practically applying AI in their digital workplaces?

In October 2017, Cearley noted at the Gartner 2017 Symposium/ITxpo in Orlando, FL that Narrow AI currently holds the most promise. Narrow AI is composed of "highly scoped machine-learning solutions that target a specific task (such as understanding language or driving a vehicle in a controlled environment) with algorithms chosen that are optimized for that task," he says. CMSWire's Dom Nicastro spoke with several experts to find some practical use cases of artificial intelligence in the digital workplace.

Real World in AI in the Digital Workplace

SAP CoPilot: Digital Assistant in the Enterprise

Chatbots and virtual assistants help us ask our phones and home devices questions. Why not bring this into the workplace setting? That's the thinking behind SAP CoPilot, a digital assistant by SAP designed to help businesses with tasks like purchasing contracts and collaborating with colleagues. Sam Yen, chief design officer at SAP, said in a video interview at SAP TechEd Barcelona they wanted to do in the enterprise what virtual assistants do for consumers. Questions like, "What's my total spend with vendor X?" can be asked via smartphone app.

SAP goal is to eliminate the need for users to manually interact with multiple work apps to get a job done. SAP CoPilot does this through a virtual human robot powered by artificial intelligence, speech recognition, natural language processing, statistical analysis and machine learning. Users can ask questions and give commands, and SAP CoPilot contextualizes their informal and unstructured speech, analyzes it and then executes actions and presents users with business objects, options and other data.

Deloitte: Machine-Learning Contract Reviews

Business consultancy Deloitte formed a partnership with machine-learning developer Kira Systems to create models that intend to quickly read thousands of complex documents, extracting and structuring textual information for better analysis.  David Schatsky, a managing director for Deloitte who analyzes emerging technology and business trends, told CMSWire, companies of all sizes need to review stacks of documents for one reason or another. They look for risks. They look at what kind of contracts they may have with suppliers or counter-parties. AI, he said, now makes it possible to do that kind of work a lot faster and more comprehensively. "It completely changes the way that kind of work is done," he says.

Headshot of David Schatsky of Deloitte
David Schatsky

Traditionally, if a company has 15,000 contracts to be reviewed, an auditor will try to get a representative sample for analysis. With AI partnerships like Deloitte-Kira Systems, natural language processing and machine learning trains the application to recognize the structure of these contracts and to enable it to pick out the key clauses, terms and conditions and the key points within a contract in an automated fashion. "So you leave behind the sort of sampling approach and you get the whole population of documents so you can get much, much better insight and you can lower the risk of missing something," Schatsky said. 

Schatsky also shared that it's possible to experiment and keep costs relatively low. Deloitte's playbook was not building AI engines from scratch, but rather a construct based on third party technology that it customized. "There's some pretty low-cost points of entry to experiment there because the big cloud providers are providing basic machine learning in the cloud," he says.

AISense: Call, Meeting Transcriptions 

We've all wanted to go back to a recorded call or meeting and find the important talking points without the hassle of trying to guess where something was said. AISense, a Silicon Valley startup founded last year, has released technology that is designed to make voice conversations accessible and searchable through its Ambient Voice Intelligence. The company last month announced it's powering transcription for the Zoom Video Communications platform and is also working to ship a consumer product available in early 2018. It raised a $10 million Series A round last month.

AISense also has in beta an application that integrates with a call-recording smartphone app. It takes recorded calls and, using artificial intelligence, transcribes those calls and curates them for users. Those leveraging the transcribed calls can search for keywords. Users can also search for terms across all their recorded calls. 

Screenshot of the integration between AISense and Zoom, which allows for recorded calls and transcription.
The integration between AISense and Zoom allows for recorded meetings and transcription.

"This is as about as practical as it gets," says Seamus McAteer, general manager of revenue and partnerships for AISense. Their technology includes automatic speech recognition, speaker identification and separation, speech-and-text sync, deep content search and natural language processing. Users
can see what was said, when and by who and output is shareable. "We believe in a world in the workplace where you can keep a record, when you want, of a conversation. Notes will go away. You can focus on what's being said. We think this can be disruptive in a very good way," says McAteer.

WalkMe: AI for Software Training

WalkMe, a digital adoption platform, offers an artificial intelligence engine that enables business software to learn about user's individual roles, habits and actions.

Athenahealth uses WalkMe to help customers (doctors and nurses) receive guidance and training on how to use a system. WalkMe integrates with Salesforce, providing personalized guidance to the user on how to create a sales opportunity in the CRM system. It also integrates with a system like Workday by including chat functionality on top, that is designed to guide a user directly to a relevant section. 

Screenshot of WalkMe's integration with WorkDay software.
WalkMe's artificial intelligence engines is designed to help users better grasp business software like Workday.

ServiceChannel: Restaurant Facilities Management Aid

Bloomin' Brands, a Tampa-based casual dining company that includes 100,000 team members and close to 1,500 restaurants, uses artificial intelligence-driven analytics to help them leverage real-time data on things like equipment, necessary repairs and operator functions. It needs to have restaurant partners, service providers and facilities teams all on the same page, Jon Ahrendt shared in their Bloomin Brands' blog post. Bloomin’ Brands integrates ServiceChannel AI-based automated facilities management technology into its facilities management processes for tracking emergencies, costs and for identifying future initiatives. 

Niles: Learning Slack Conversations

Niles is a Slack add-on that listens and records conversations that happen within the collabortion platform. PPC Protect is an organization that uses Niles often. Every time someone sends a message, it learns, according to Maria Hugh, systems manager, PPC Protect. Users can ask Niles questions, and using AI it will respond with an answer. "What products do we sell? What sizes? How much do we charge? Who's in charge of this department?" Everything is stored in a database and can be recalled in the future.  If Niles fails at an answer, users can provide him the right answer so Niles is always up to date. The more you interact and use him the more he learns and gets things right. 

Screenshot of Niles, a Slack add-on, getting into a conversation on a Slack channel.
Niles, the Slack add-on, in action. Here, he provides the wrong answer by providing the wrong link. AI is always learning.

MESH, Not Rocket Science: Branded Bots

MESH, a Baton Rouge-based strategic marketing firm, and Not Rocket Science (NRS), a Covington, Louisiana-based software cognitive business solution development firm, have built "Branded Bots," artificial intelligence applications that aim to deliver brand personality through various channels and platforms such as Alexa, Cortana, Siri, IBM Watson, mobile apps, websites and social media. "Our goal is to take a job, order or manufacture from pre-production and proposal through billing with AI technology." said Taylor Bennett, CEO and Founder of MESH. The example they offer is a bot they've developed that focuses on software integration and operations, linking together project management, time tracking and CRM systems.

One user of their technology is Milo Ag, an agricultural company that provides producers access to cognitive solutions, natural language interfaces and blockchain. Milo is their cognitive natural language personal assistant who is an informed expert on trading, options, regulations, compliance, transportation, crop insurance, financing, agronomy, machinery, farm programs, weather, news and more. Milo's goal is to help guide producers and their employees to make the best decisions.

Screenshot of Milo AG, which uses AI technology from Rocket Science & MESH to create custom AI /bots for their businesses.
Milo Ag uses AI tech from Rocket Science & MESH to create custom AI/bots for its business.

Acculation, Inc.: AI Meets Social Media

Werner G. Krebs, Ph.D., CEO of Acculation, a data science software company and consultancy, told CMSWire his company uses data-driven processes to make decisions about content for social media and elsewhere. Data can be used to make social media content decisions and AI and algorithms can actually create the content by themselves, according to Krebs.