Mitch Coopet believes that artificial intelligence is more likely to build killer sales teams than killer robots. Two years ago, Coopet launched Aftercode, a startup that uses AI to analyze the content and information of sales calls. Prior to founding Aftercode, Coopet was the co-founder and chief technology officer of Minneapolis tech company Code42. Aftercode raised just over $2 million in seeding funding last year and debuted its first product, Rambl.

Minne Inno sat down with Coopet to discuss Aftercode’s growth and some of the potentials and pitfalls of artificial intelligence.

Note: This conversation has been edited for brevity and clarity. 

Aftercode CEO Mitch Coopet. (Photo by Nancy Kuehn, Minneapolis/St. Paul Business Journal)

Minne Inno: Where did the idea for Rambl come from?

Mitch Coopet: We had a sales team of over 100 at Code42. I thought, ‘It would be so great if I could query all the calls these people are making. That would be incredibly valuable.’ But that was impossible at the time. I felt there was this huge amount of data out there that we weren’t capturing. What are the people who aren’t buying our product saying? I’d love to know that. The number one problem in sales right now is that reps don’t log their calls in their CRM systems. Today, reps often take 20 minutes to fill out something like that. With Rambl, it’s one click and part of the calling process. Imagine if I can save a rep 20 minutes per call. That’s not even AI. That’s just creating a better phone system.

MI: Where does the AI come in to the process?

MC: We work with sales managers to develop qualification criteria, sometimes called a playbook, and Rambl examines the contents of the call and assigns it a score, which tells managers how well their reps are adhering to the criteria they set. A lot of times people get really caught up in ‘Oooo AI is going to change the world and make scary robots.’ How about we just try to detect a voicemail better? We can change the way people make phone calls. That’s what I’m really excited about. 

“Selling a sales tool to salespeople is the best thing in the world because you’re always learning how to be better at selling”

MI: What’s it been like trying to sell a sales platform to salespeople? 

MC: I’m not a salesperson or sales leader, but I’ve been a founder and I’ve had to sell my ideas. In a strange way, founders are the ultimate sales people. I’ve worked with sales teams and I have a huge admiration for what they have to do. All these things we’re helping sales teams do with Aftercode, we’re doing as a company too. Selling a sales tool to a sales team is the best thing in the world because you’re always learning how to be better at selling.

MI: How did you first become interested in AI? 

MC: I was hearing a lot of buzz about deep learning, machine learning and data science. At the time I didn’t give it too much attention, because it’s very statistical. I’m an engineer and programmer, which is a very different, data-focused type of mindset. My interest came from when deep learning was able to start recognizing patterns in unstructured data that weren’t recognizable before. My whole thesis with building value with technology starts with sources of data that are currently raw. If we have the ability to refine that, we’ll be able to get a lot of value out of it.

MI: What are some challenges you’ve seen or encountered with AI?

MC: I think the biggest problem with AI at this point in time is probability. As humans we seek certainty, and AI can’t provide that right now. Think about Siri. She’s not able to answer most of the questions you ask her. And that’s the problem. People don’t want to deal with probability. They want certainty. The reason they’re paying money for tech solutions is that they want to eliminate uncertainty or risk. For an AI system to be accepted you have to achieve perfection, which is impossible. Really, it’s more of a problem with user experience and expectations than AI itself.

MI: So how does Aftercode address that issue?

MC: Our entire approach is to narrow the scope so we can be right a lot of the time. I like to say that we can’t find the needle in the haystack for you, but we can burn a lot of hay and make it easier for you to find it for yourself. 

MI: Are there any AI trends in particular you’re interested in?

MC: I think deepfakes are fascinating. I think now that people have realized what they can create, it brings a whole new meaning to photoshopping. It’s scary and it’s interesting. This ability to create fake videos and voices is going to be one of the biggest challenges of the digital age. But it has interesting and compelling applications in entertainment. 

MI: Since this is your second company, any lessons you’d share with first-time founders?

MC: You are the only one who knows your business. Own your idea. Own the domain. No one can tell you that they’re the expert of your business. If they were, they’d be building your business. So much of the authority to influence an early founder is given to their advisors. You are the one taking the risk. Don’t let someone else make those important decisions for you. Believe in yourself and don’t just give control away to people because they have a title or express an opinion that sounds very fanciful. They don’t see what you see. And it’s ok to disagree.