• by Laurie Sullivan , Staff Writer @lauriesullivan, August 3, 2017

Optimove CTO Tal Kedar's best ideas come to him while in the shower. Some uncertainty about the future of technology and ad tech keeps him thinking, but it doesn't keep him up at night. Instead, he tries to create an environment within the company that allows people to grow. 

"When you're singing you sound best in the shower," he said. "Similarly, your ideas are best in the shower. At least for me."

As CTO, Kedar will focus on translating the needs of the business into technological solutions. Speaking both languages and being able to translate one to the other is at the heart of what CTOs do, he said.

Kedar, who joined Optimove about four months ago, specializes in artificial intelligence (AI). He says the ability to match the approach to the business problem is a huge part of how AI really works. Sometimes marketers will use the buzzword because it makes sense, but the algorithm used has nothing to do with AI, he said. Companies that have a good understanding of the business problem and their customers, as well as available approaches in AI, can do it well.

He acknowledges that it takes time to learn.

When I asked Kedar why it sometimes seems the machines are teaching engineers how to teach the machines, he said that with the emergence of deep learning and neural networks came a new way of trying to answer questions.

Engineers don't know how to use the technology completely, because the methods are so new. Trying new things help people learn. In some cases, it looks like the machine is telling you how to teach it because you're allowing the machine to run many architectures and models. So when you look back at the most successful results, it looks as if the machine told you the best way to teach the machine or algorithm image recognition, for example.

He said neural networks maybe a good approach to optimize image recognition, but not the best approach to optimize marketing campaigns. Understanding the type of AI approaches and algorithms to use for specific business problems has become very important.

Optimove doesn't treat AI as a buzzword, he said, while talking about his time at Sears where he served as the director of emerging technologies.

Kedar also used AI at Sears. His team spearheaded search engine marketing optimization on Google and Bing, as well as product recommendation. The amount of data presented an amazing opportunity, but it also required known AI algorithms to make predictions. There was so much variety in the data because someone might buy a lawn mower, pair of socks, and refrigerator in the same order.