Role Models in AI: Omar U. Florez
As a Senior Research Manager in AI at Capital One Bank, Omar U. Florez continues to learn and code the latest AI algorithms to improve the company’s capabilities.
Born and raised in Peru, Omar grew up with a fascination for computers and robots. This curiosity led him into the field of computer science and later into AI development. He is passionate about AI and wants to pass on his knowledge to Latinos through the development of an open source Spanish-English AI dictionary with the help of Capital One. He is also the Research Committee Lead for LatinX in AI Coalition, where he works on developing data and research to accurately represent Latin American culture.
When he is not reading about AI or researching the latest findings in the field, Omar is thinking about his next destination to scuba dive.
Learn more about what sparked his passion for AI, his work on creating more diversity in the field, and the role models that inspire him.
We interviewed Omar as part of AI4ALL’s Role Models in AI series, where we feature the perspectives of people working in AI in a variety of ways. Check back here for new interviews.
As told to Eunice Poon of AI4ALL by Omar U. Florez
EP: As a Senior Research Manager in AI at Capital One, what types of projects are you responsible for and how do you use AI in your work?
I work to implement new or recent techniques or algorithms with the goal of providing natural conversations between humans and computers. Basically, I use natural language understanding to create a new experience for our customers through our virtual assistant, Eno. We want Eno to communicate naturally, to provide a good outreach experience to our customers.
Every week is busy but very interesting. My week is usually filled with lots of experiments to test new ideas or to actually see if the algorithms are working or not. The main goal is always to impact the main products that we have in the company and make them better and more efficient for our customers and our associates.
I was also involved in creating an AI dictionary at Capital One. The dictionary translates English to Spanish to enable people who are Spanish native speakers to learn AI in their native tongue. The support from Capital One in making this a reality was huge because it is a great tool that opens up AI education to young Spanish speakers to pursue a career in AI, and creates more diversity. This dictionary is shared on GitHub and is freely available to anyone who wants to use it or add to it.
EP: You have a great deal of work experience as a computer scientist. What sparked your interest in CS in the first place and how did you learn about AI?
When I was a kid, I was always fascinated with computers and how they worked. I wanted to understand what made computers different than robots because I felt like computers were little robots in some ways. So in college, I started to study computer science to see if I could actually turn a computer into a robot.
During my studies, I learned that computers have all the computational power to be a robot, it just uses a different kind of model, engine, or sensor. I was also challenged by one of my professors who told me that computers can in fact predict the future. I didn’t know that was possible at the time. This was when I realized that AI was something even more specific to what I was interested in as a child. The power to predict the future based on evidence and hypothesis is fascinating, so I decided to continue studying AI to expand my knowledge. Today, even after my PhD, I am still learning about AI every day.
EP: What is the perception of AI like in Peru? Are there a lot of opportunities in the field? Overall, do you think people hold positive views towards AI?
In Peru and Latin America in general, there is interest in AI from students, universities, and companies. I think right now, there is a lot of emphasis on data science and increasing interest in AI and automation.
To be honest, I am a little worried about Peru’s strategy towards AI. For example, in Peru and Latin America, there are a lot of manual jobs that could eventually become obsolete or changed by the development of AI technology. I’m worried that we don’t have a clear national strategy, knowing that we could be one of the most impacted regions in the world when it comes to AI development.
I think as people in the AI industry, we have to start encouraging governments to start thinking about maximizing the long term gains of these techniques and opportunities, and discovering how to benefit society with the technology. We need to start thinking about AI’s impact on our countries, 10 or 20 years from now and plan ahead.
EP: What are some of the important things people should be doing to create a positive future for AI? And how are you creating a more positive future through your involvement in LatinX in AI?
Creating a more positive future for AI centers around two main topics: education and ethics.
I think that everyone should have the same level of opportunity when it comes to AI education independent of race, country, age, or gender. We should all have access to the latest discoveries in AI and the most updated data sets. This access I believe can create more diversity, because everyone will have an opportunity to learn.
I also believe that scientists and engineers have an obligation to treat people equally and ethically.
So creating a positive future in AI requires these same scientists and engineers to care about creating algorithms that treat everyone equally, and benefit everyone equitably.
I lead the research team for LatinX in AI, and our goal is to create harmony between AI and Latino culture. In order to do that, we start by raising awareness about data inaccuracy. Some data sets or algorithms may not accurately present LatinX culture, so we want to discuss that problem and encourage the creation of new data sets and research projects and make them open source. Through this, we believe that other scientists and engineers can use our new data sets, for example, to develop research that can fairly represent Latinos and Latinas.
LatinX in AI also had the opportunity to partner with NeurIPS last year, allowing Latinos and Latinas to travel from different parts of the world to the conference. For the very first time, we had so many individuals speaking Spanish at the event. It was a really amazing thing to be involved in. This year, we are organizing the same workshop for ICML.
EP: Who were your role models growing up? Do you have any role models now?
When I was a kid, my role models were my parents. They turned down the opportunity to go to school because they took care of our family instead. They provided for us and I am very grateful for that.
Right now, in terms of research and innovation in AI, my role models are people like Andrew Ng, the co-founder of Coursera and GoogleBrain, and Timnit Gebru, the technical co-lead of the ethical AI team at Google. I had the opportunity to meet Timnit, and she is an amazing person. Not only is she technically skilled, she is also one of the humblest people I’ve ever met. When you talk to her, you feel like you’re talking to a friend and that’s a great lesson by itself. In the case of Andrew Ng, you can see a person who is very technical, but also has a passion to share his knowledge with everyone. For me, this combination of sharing knowledge and opportunities, and also being approachable and humble, is an amazing combination.
EP: What advice do you have for young people who are interested in AI who might just be starting their career or academic journeys?
The biggest advice I have for young people is to stay curious. Stay passionate about learning. Learn everything you can, because when you start working it is a very dynamic environment and there are always new things to learn. Get excited about reading and taking in new information, delve deeply into topics. If you stay curious and see learning as fun, you will enjoy the work and continue growing in the field.
About Omar
Omar U. Florez is a Senior Research Manager at Capital One working on enabling natural conversations between humans and devices. He is also the research committee lead of LatinX in AI, a coalition that creates harmony between Latino culture in the US and the recent advances in Artificial Intelligence. Before, he was a Machine Learning Researcher at Intel Labs (California) focused on teaching computers to understand user context by discovering patterns in people’s routines from visual and textual data. He uses Deep Learning for accurate predictions and Bayesian models for discovering interpretable hypothesis. Flórez is also a recipient of the IBM Research Innovation Award in large scale analytics.
