AI and Business Tech: Intelligent Conversations with Chatbots
Artificial intelligence (AI) is getting a lot of attention these days from the media and press, from Hollywood, and even from the business sector. You may have also heard about AI being used in business software. AI-driven software can include a wide range of AI mechanics, including text and image processing, financial modeling, sales forecasting, Internet of Things, machine learning, or recommender systems (to name a few).
But what does this all mean? How exactly can AI help your business? Well, that’s not something we can answer in just one blogpost, so we’ve decided to launch a series on AI and Business Tech.
In this first blogpost, we’re going to discuss one of the most common AI mechanics found in AI-driven business software: natural language processing. But before we get to that, let’s first talk about what we mean by intelligence.
True Intelligence vs. Artificial Intelligence
When you ask Siri what the weather is like today, she exhibits a very basic form of single-layer artificial intelligence. First, she converts your speech to text, parses the text to identify a question, locates the best possible answer in an available database, and then finally delivers that answer to you.
When you ask a human what the weather is like today, they exhibit true multi-layer intelligence. You may get the same answer that Siri gave you, but you might also get additional facts in context. For example, your spouse may know that you were thinking of going for a run. They might say, “It’s a little rainy out right now, but it’s supposed to clear up later if you want to go for a night jog!” Current AI technologies don’t understand context very well, and therefore can’t deliver this kind of information.
What AI does understand very well is data, patterns, and statistics. AI can look at different data points and give you suggestions based on that data. For example, most evenings after a day at the office, my smartphone GPS application automatically tells me how long it will take to get home. But one summer evening, after I visited the marina where I keep my boat a few days in a row after work, my phone told me how long it would take to get to the marina instead.
I never manually searched for any of these destinations, but my phone recognized data changes in my travel patterns and applied a statistical model to predict where I will probably go at a certain time of day. This may not be true intelligence, but it is an example of well-designed AI.
Natural Language Processing and Chatbots
Natural language processing (NLP) is the mechanic behind your exchange with Siri about the weather. NLP is exactly what it sounds like – the analysis of language (audio clips or written text) for the purpose of gathering or providing information. Even though current AI technologies can’t come close to the true natural language intelligence of humans, they can apply NLP mechanics in interesting ways. A good example of NLP mechanics in business technologies is customer service automation.
Many businesses use an AI platform called DigitalGenius to automate first-line customer support requests. Consider this real-life question submitted by an airline traveler via live chat on the airline website:
“Hey guys, I left my phone on the plane yesterday. How can I get it back ASAP?”
A DigitalGenius chatbot applies NLP mechanics to the question and analyzes key phrases and syntax to understand the question better. The chatbot first determines that the question was submitted after travel (and therefore by a customer, not a prospect) regarding the lost item. It also recognizes that the issue raised is high-priority but the customer is communicating with neutral sentiment. So, the customer is likely looking for a quick answer but is not angry or upset enough to need to talk to a human. The chatbot then replies:
“All found items are handed over to the airport lost and found department.”
The customer got their quick reply and was able to get the phone back at the airport. The airline did not need to waste time or money resources to engage with a human and the customer did not have to wait on hold for a human to tell them what a bot was able to tell them in seconds.
How Else Can This Help Your Business?
Simple: through integrations with other tools – AI or not.
If the chatbot from above was on your website, it could save you and your customers time and effort. But it doesn’t stop there. Data on these conversations is saved in a data repository or data lake.
Let’s say your company uses a software tool for quality control on the factory floor. The quality control tool determines that a flaw in factory equipment affected some of the products shipped last quarter. That information goes into the data lake. If this tool is integrated with your AI chatbot, the chatbot can take this information from the data lake and forward a customer to a customer service representative if they happen to ask about those defective products.
Now let’s take this a step further. A workflow assistance tool within your CRM might suggest actions or send notifications to users. The tool might flag any opportunities, contacts, or accounts associated with those defective products and pour that information into the data lake as well. Any time the chatbot recognizes that they’re conversing with a customer associated with a defective product – even if they’re not talking about that product – the chatbot can inquire about the product or redirect them to their account executive. This helps salespeople nip potential problems in the bud and build trust with their customers.
What About Other AI Mechanics?
We haven’t even scratched the surface of how different AI mechanics drive each other to perform complex functions within business software. We have a long way to go before we develop true intelligence within software, but AI-driven technology is already here!
Stay in the Loop
Subscribe to get all our latest content sent directly to your inbox!