Businesses (and their investors) are atwitter about AI in the enterprise these days. Yet with all the buzz, there is a disproportionate focus on those use cases which support sales and marketing. Tractica’s research finds application for AI in every function of the enterprise. This is in part due to the diverse capabilities that fall under the AI umbrella. For example, natural language processing (NLP) supports language processing whereas computer vision supports understanding of image or videos. These are but two of a number of technology subsets supporting enterprise use cases for AI.

Access this post where it was originally published here


Potential applications for AI in the enterprise run the gamut, including sales, marketing, customer service, operations, legal, finance, human resource, IT, supply chain, merchandising, logistics, research & development, and far beyond. Despite a frenzy of attention on chatbots, companies must think broader. This is particularly challenging because many successful applications for AI today are quite narrow, meaning they support very specific business pains with incremental advancement. Although transformation is sexy, incremental automations can make significant impacts to productivity and efficiency. Let’s look at three examples.

There are AI applications that apply to specific functions, such as HR.

AI is now being applied to save time, energy, and money during the talent sourcing and recruitment processes. Models mine large data sets, third-party job sites, and social media to source candidates with higher likelihood of interest and hiring potential. These models help sort multiple resumes, mine text for specific needs, prioritize, and surface candidates. some systems are also using ML to drive employee engagement. BetterWorks, for example, focuses on using AI to ease the employee-manager feedback loop. It does this by building work profiles, which it calls “Work Graphs” based on data integrations across Google Apps, Office 365, Salesforce, JIRA, email, and Slack, then track employees’ goal progress, alignment, comments, cheers, budgets, cross-functional collaboration, etc. to inform employee engagement strategies. Specifically, they use machine learning to prompt contextually appropriate feedback, recognition, council, questions, and learn from employees’ preferred channels, time of day, etc.

There are AI applications that apply across functions.

As an extension of image recognition and analysis, AI is now also being used by organizations to aid in audio and video mining. In business, audio and video are useful media for storytelling, brand awareness, business development, customer support, and education. Until recently, such efforts, whether owned, paid, or earned, have been difficult, if not impossible to mine for insights. Various AI technologies can now help organizations begin to leverage these insights at scale. In a marketing or market analysis context, speech and voice recognition can be mined for specific moments, such as a user posting a video about a product. In a call center context, companies like DeepGram and demonstrate how AI can be used to transcribe, identify keywords, and mine phone calls, video footage, or online media. Deep learning can also be applied here for auto-generated speech-to-text transcription.

There are AI applications that help bring functions together.

Most companies, particularly large multi-national corporations, struggle to collaborate and coordinate as a unified organization. As the amount of data flowing into and across organizations grows more and more massive, the problem is not just one of content distribution, but of the time it takes to comprehensively identify and organize insights that are useful and consumable. AI is now a tool well suited for automated report generation. AI offers the potential to collate reports far more rapidly and easily configurable than humans. By supporting tasks like data sourcing, data interpretation, data analytics, and narrative commentary, automated report generation tools can surface relevant metrics, tables and charts, and generate multiple paragraphs of narrative. Companies like Narrative Science, Arria, Genpact, and many others are demonstrating various configurations of NLP, machine and deep learning to deliver reports on everything from advertising performance, to sales, to employee satisfaction, competitive intelligence, and beyond. 

These examples are just three of some 215 use cases Tractica has identified across markets, but they offer a glimpse into the utility of AI technologies commercially available today. With that said, it’s essential companies approach AI applications with caution, limited expectations, and most importantly, good data.