How to Use AI to Increase Business and Make Customers Happy

A teacher of artificial intelligence engineering at Columbia and NYU is adamant: Taking a human approach with your customers can set your company apart.

Andrei Papancea


How to Use AI to Increase Business and Make Customers Happy
See the original post on

As the co-founder of NLX, a conversational Artificial Intelligence company, I have a people-focused/solution-oriented approach.

There’s a tendency to glorify AI and make it sound like it’s going to take over the world.  NLX was born in 2018 with the mission of helping companies transform customer contact into personalized self-service.

Along the way, I’ve learned that means to be a people-first leader, to always take the human approach with your products and customers.

Here’s what that means...

Designing people-centric products
When it comes to a product or service-related problem, the issue is not always your ability to solve it, but the utility of a fix.

We’re organizing and simplifying AI. We want to make it easy for non-technical clients to be able to take charge and build applications their users truly need. People like feeling empowered.

Furthermore, we’ve made the conscious choice to not build our own conversational AI model, because there are multi-trillion-dollar companies heavily invested in doing just that. Instead, we ask ourselves: “How do you manage content in tens of languages while streamlining integration for better automation?”. 

What drove products five years ago is coming to pass and putting technologies up against each other may not provide solutions. This is why, for instance, we don’t lock in unhappy customers in multi-year contracts and use pay-as-you-go pricing instead of the tiered pricing that is the norm.

Stick to human-oriented design and solutions. It will help you focus on what’s truly important as you build your products and that will set you apart.

Creating value-driven cultures
As I was reflecting on the past jobs, I realized that people left for three reasons: A) They weren’t paid well, B) They didn’t feel respected, and C) They didn’t have engaging work to do.

In building NLX’s culture, we made it a point to not make those mistakes and the results have been extraordinary. To an external investor, perhaps saving a few thousand dollars by letting a resource leave might sound right, but as a founder, only you can know the real cost of losing a great colleague.

Your team imbibing and embracing the values you encourage will carry the day. Those values will extend to your customers through the experiences that are created for them. 

Building customer-oriented companies
Instead of thinking about what technology you can offer your customers today, focus on their experience and work backward from the ideal solution. Doing this allows you to break the pattern of what the status quo provides. It helps you truly innovate on behalf of the end-user, your customer, your platform and your industry.

What we’ve learned is that customers and companies want solutions. And when they’re struggling with a problem, telling them about a new tool or the latest transcription feature may not be the same thing as actually giving them a solution.

We found that one of the top three reasons people call call centers is to reset passwords. Everyone assumes this is easy to do on a website and, yet, customers always struggle because it's hard to automate resets over voice instruction, (i.e., the customer needs to go “uppercase ‘A', lowercase ‘b’, zero ‘0’," etc.).

It’s the same with automating airline bookings. My last name is Papancea and, like me, there are millions of people out there with non-regular-sounding last names. Going with the status quo of the tech today in such cases does not work. It’s hard to capture last names and booking codes such that they are transcribed correctly. In building solutions for such use cases, we decide to work backward. If that mean building multi-modal capabilities to transcend channels and go beyond the boundaries of traditional automation? We do it.

If you pay attention to your metrics and listen to your customers, they will tell you what they really want.  Visualize an ideal, or a delightful way in which a user would like a problem solved, and then build them an automated solution.

It’s easy to get excited when you get a big customer and sad when opportunities don’t pan out. The emotional oscillation can get unhealthy, so know your value and stay steady. Celebrate the smallest milestones, but also cherish things that don’t work out. They always show you new paths to explore.

Andrei Papancea

Andrei is our CEO and swiss-army knife for all things natural language-related.

He built the Natural Language Understanding platform for American Express, processing millions of conversations across AmEx’s main servicing channels.

As Director of Engineering, he deployed AWS across the business units of Argo Group, a publicly traded US company, and successfully passed the implementation through a technical audit (30+ AWS accounts managed).

He teaches graduate lectures on Cloud Computing and Big Data at Columbia University.

He holds a M.S. in Computer Science from Columbia University.