Artificial Intelligence that keeps humans in the loop: How, why Swivl is turning heads

Editor’s note: This is the latest in YourLocalStudio.com’s “UpTech” series focusing on Artificial Intelligence. Today, Alexander Ferguson, CEO of YourLocalStudio, takes another deep dive into how AI can be used in business.

UpTech

Artificial intelligence, machine learning. These emerging technologies are changing the way we live, work, and do business in the world for the better. How is AI actually being applied in business today though? In this episode of “UpTech Report,” I interview Mason Levy with Swivl, whose aim is to simplify AI training by turning customer data into machine learning-ready models, or as they call it, a human-in-the-loop AI concept.

The interview:

So, in 30 seconds, define how does the swivl product help a business, and who can best utilize it right now?

Ultimately we typically sell into customer success teams and help automate the customer interactions through natural language understanding, similar to chatbots and voice bots on the phone.

Can you explain good use cases, and who might best use the platform? What’s a good use case for customers, and what is often the first thing that swivl really can help them with?
We are very good at onboarding onto a SaaS platform or creating eCommerce product recommendation engines. These are really great starting points to get started. It allows companies to get their hands dirty with something that immediately provides value. We can then we can branch off into other avenues of automation with that business as they learn the system a little bit better.

How would you say that your product is different from other ones in the market?

I think our core competency is in the feedback loop. Many products that are out in the market allow for a model to be built and for you to start to interact with your customers. But the system is not learning, its not getting smarter, We try and really double and triple down on that feedback loop. So when the model isn’t competent or a conversation fails, it escalates to a human, who trains it, and the platform learns from that human being interaction. Ultimately you see the success of that model as it continue to get smarter over time.

Mason Levy from his LinkedIn site.

Technology’s awesome when it works, and when it doesn’t, then people are like, “This is an annoying, stupid machine.” And that’s why I find interesting that the human-in-a-loop concept can help with this kind of carryover. But let’s talk more about the actual technology behind your platform. How are you utilizing artificial intelligence and machine learning and other technologies?

Yeah, ultimately we sit on top of natural language understanding models in order to determine intents within user examples or utterances, as you might want to call them. We also identify entities. We then rout them properly to automate the response, the business process, or escalate the conversation to the appropriate team. So in some cases, we’re not actually automating a response, we’re just identifying the intent and figuring out should this go to the finance department or should it go to the customer success team?

For those that are trying to build their own product out there using natural language processing and AI, do you have any suggestions that you would give to them on issues they may run into-choosing or creating their own based on this type of technology?

I think the hardest part is getting started. I think a lot of entrepreneurs, especially in the natural language area, will spend a lot of time and expense trying to build a system. Instead, the best way to get feedback is to get out there witha platform like swivl, get live action, and then have a system that allows for that feedback loop to be closed – to gather feedback, and improve the system.

Again, a platform like swivl really helps with that. We can plug in to a lot of different channels, whether that be Slack, Microsoft Teams, WhatsApp, Skype, etc. I think the biggest hurdle is don’t be scared to put it out in the world and let somebody give you some feedback.

How many companies are you guys helping right now with your product?

We have 32 customers on the platform. Everything from a one person team up to 100 people within a single large enterprise who are using it. We’re really excited about where we’re at right now.

Where do you see yourself in five years, where is swivl going to be?

Yeah, we really look to be the “Uber of Data Labeling.” I would love to walk into a coffee shop for a meeting and see somebody sitting down, waiting for somebody else to meet with them, using the swivl application to label data for some large enterprise company down the road.

So if somebody wanted to come and work with you, with varying sizes of company, what are they looking at then, expense-wise at being able to utilize your platform?

Yeah, the beauty of our system is that you can get started for free. We offer Free Live Chat, this allows for the data to come in and become part of our training queue. As soon as you want to turn on the training, our lab’s model starts at about $250, and then as we scale with volume as well as the number of users. Our enterprise models start around $5,000 a month. So we really have a price point that fits just about everybody’s needs.

Awesome, thank you again for joining us today. For more information, where can people go and what’s the next step that you would recommend that they take?

Look at us up at tryswivl.com, as I mentioned, you can go in and sign up for Live Chat to get into the tool and started today. Or reach out to me at mason@tryswivl.com. I’m happy to do a full demo, show you some of the use cases that have been successful and discuss ROI that our customers are already seeing using the product. I look forward to chatting with anybody that’s interested.

That was just one example of applied AI. Have you seen a company utilizing AI or machine learning to transform the way we all live, work, and do business? Contact us at YourLocalStudio.com