Sofia, our AI-personal benefits assistant, is one-of-a-kind. And as smart as she is, she doesn’t grow and learn alone.

There is a team of unique individuals behind the scenes teaching her new skills and helping her keep her knowledge base and innovative methods cutting-edge.

I was able to sit down with Yanan Jia, Sofia’s Machine Learning Research Scientist to showcase her projects and role in developing Sofia. Here are some highlights from the conversation. **the text below has been edited for clarity and length.

What has been one of your favorite projects working with Sofia?

The sentiment analysis project. Sentiment analysis allows AI to understand a more natural voice for better recognition and is crucial for the advancement of AI. I actually started listening to calls in the service center to understand the different ways employees could ask a question or, understanding the nuances of different emotions, like anger. Studying these human language interactions allowed us to help make the interaction between humans and AI more intelligent.

Our work on sentiment analysis actually got published by Cornell University.

How do you see AI evolving?

AI will continue to evolve. A new form of joy and productivity is human and AI collaboration. For example, Sofia helps free up a lot of time for our representatives. She is a valuable partner and allows our service team members to handle the more detailed or complex calls that need more time and attention. 

Of course, there are always questions that Sofia cannot answer. We really need a human to mind the conversation and teach Sofia how to handle those questions. Human skills and AI abilities will become the new generation of workers. AI is not a threat, but a partnership.

What’s your favorite part of your job?

I received my degree in 2016 and this is my first full-time job. I’ve been with Businessolver for 4 years. I really like the working environment. When I went to school, I learned a lot – but it’s been great to be on Sony’s team where I get to explore and learn about new technologies and theories we can harness and use to develop Sofia even further. Machine learning has new techniques every year, sometimes even every month. By understanding the latest cutting edge data models, we can continue to increase Sofia’s knowledge and ability to understand human speech patterns so she can continue to evolve and get smarter every year.

Watch the full interview below. And, if you would like to learn more about Sofia, check out our infographic. 

And, if you're curious about the other rock stars on the Sofia team, check out our recent interview with Lexi, our Computational Linguist.