As a conversation engineer at Autodesk, Yizel Vizcarra enhances and optimizes the dialogue of the Autodesk Virtual Assistant (AVA). Recently, she was one of the leads on a groundbreaking project to add voice and motion to AVA, adding a video chat feature.
Her varied interests led her to fall in love with the interdisciplinary field of cognitive science. After pursuing a research opportunity and working as an Autodesk intern, she impressed a VP at the company with her insight into the value of playfulness and cognitive diversity in tech. This encounter led her back to Autodesk fulltime as an engineer.
Yizel is a first-generation woman of color who cares deeply about multidisciplinary teams and being transparent about uses of AI. Learn more about what her day-to-day looks like, how her role models teach her to be brave, and her advice for those looking to enter AI.
We interviewed Yizel 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 weekly for new interviews.
As told to Nicole Halmi of AI4ALL by Yizel Vizcarra; edited by Panchami Bhat
NH: You’re a Conversation Engineer at Autodesk. What does a typical week at work look like for you? What kind of responsibilities do you have and what kind of projects are you working on?
YV: A conversation engineer is responsible for the dialogue and interaction between a human and a customer support bot. I work on development of the bot to enhance its features based on business context. I teach it to understand questions and really deliver customer answers.
I really enjoy looking at how customers are engaging with our virtual assistant. I take a look at that to get more perspective into what our customers are really asking for, what our customers need or don’t need.
I’m currently working on a project that adds a voice, face, and motion to our avatar. We collaborate with a company in New Zealand called Soul Machines and two weeks ago we launched our first virtual animated assistant. It’s built off of our text interaction and allows us to offer a video and audio chat option. I was the lead in the development of that voice and the conversation and the UI (user interface) that essentially walks the customer through their interaction with the virtual assistant.
Learn more about Yizel’s day-to-day in this “Day In The Life with Autodesk’s Yizel Vizcarra” video.
Your work strikes me as quite multidisciplinary — you need to have empathy and understanding for the people using your service, you need to be a strong writer, and you need to have technical and analytical skills. This is not what people necessarily think of when they think of jobs in AI. What has your experience been like in the field?
At a recent conference I attended the keynote speaker was talking about how we should push to have more cognitively diverse teams, or people from different disciplines.
For a job like mine, where you really have to understand the back-end all the way to the front-end and real interactions with the customers, you need both technical and analytical skills to fully see the issues and deliver quality answers. My team is very interdisciplinary. We have people with philosophy, computer science, and linguistics backgrounds. I have a cognitive science background, so that’s interdisciplinary in itself.
A combination of all of these fields come together and we account for each other’s blind spots.
We also all have a common language: we all understand the technical aspect of the bot. We use that common language to approach the issue from a very technical perspective, without compromising innovation.
We need empathy and understanding for the people using our bot, which means that we need to engage with what our customers are saying, and build the best product based on that. I feel this is a dialogue that’s lacking in AI because it’s sometimes extremely technical. Even so, because AI impacts humans, we need to have multidisciplinary teams working on projects like this.
How did you decide to get a degree in cognitive science? Were you interested in the field at a young age, or did you discover it in college? And how did you come to work as a conversation engineer?
I had no idea what cognitive science was until I was in college. Someone I knew said he was going to enroll in a cognitive science degree. I thought he was cute so I went back to my room and Googled what cognitive science was. After, I forgot about the boy and fell in love with the major.
I enrolled in an intro class and after the first day I knew that cognitive science is what I wanted to do. I’ve always been someone who likes many different things, but couldn’t find a focus. I felt that this major really encompassed that by tying together a blanket of similar topics related to how the brain works.
I didn’t really enjoy tech in college — computer science classes were more difficult for me than I would have liked. I felt like I had to catch up with everyone and learn what everyone else already knew, because I didn’t have the same math background that most people have in school. Even though we had support, and had good math teachers, I felt that our high school didn’t have a lot of resources that we needed.
I feel that I was very lucky in my career path, in the sense that the opportunities I chased led me to end up where I am now. I have many varied interests, and had no idea what I was going to do after college. I interned at a research and development lab for Kimberly Clark in Wisconsin.
Then, I learned about Autodesk. I really liked their diversity efforts, so I applied for an internship. After going through a few rounds of interviews, I ended up getting the position as a knowledge engineer intern at Autodesk. In that role, followed pain points to try and build a game plan for how we can help the data community at Autodesk.
I mentioned to someone that I was interested in the human perspective and artificial intelligence. He was very invested in putting me in a place of my interest, which is something I love about Autodesk. There was a project happening where we were trying to build a virtual assistant. This was a space where AI, psychology, and the brain, really came together. He pushed me onto this project, which has been a very exploratory experience.
The manager at the time wanted to get people in the company to understand building a virtual assistant is complex. To that end, I met with one of the VPs as an intern and I played a game with him. I had him organize flash cards into categories. He organized them incorrectly, and that was my point. He laughed when I had him play the game, but later came to me to say he still talks about the meeting. He said that I delivered the message that because our brains are wired in different ways, this is a multidisciplinary effort — and I did it in an engaging, interactive way.
Because of that meeting, he really pushed for recruiting me. That’s how I onboarded into this conversation engineer AI role.
As we continue to solution things in AI, I see that different perspectives are really valued, particularly as the field is emerging and requires a lot of innovation.
What are some of the things people should be doing now to create a positive and inclusive future for AI?
People should be very transparent about AI and have more dialogue about what that means for society. Doing so makes it open for people who are passionate about making life better for humanity and people who value the technological perspective. Being transparent about why people want to automate, what the purpose of automation is, what using AI could do, and how it could be beneficial instead of something so scary is important.
We should also actively be encouraging others. I come from two immigrant parents and I grew up in the hood of LA, in Inglewood. I had people believe in me when I didn’t believe in myself. For example, the VP who helped bring me to Autodesk. I was really nervous when I introduced that activity to him, and he took the time to tell me what that meant to him and why it mattered. Even if they’re simple words, it’s important to encourage people from minority backgrounds to believe in themselves when you believe in them.
Who were your role models growing up? Do you have any role models now?
I can be a very quiet person, so my role models are people who teach me to be fierce and to really put myself out there or force me into uncomfortable situations.
My dad was and is one of my biggest role models. When my mom opted for me to not play soccer, he advocated for me, saying I should be able to play. He was always pushed me beyond the boundaries that were set just because I was a woman of color in the US, and encouraged me to go against the status quo. He also is a very brave person for leaving Mexico at a young age and trying to make a better life for himself and the family. I always call him my Superman.
One of my directors is my role model today. She is a woman and leads a technical group. When she goes into a room, people listen to what she has to say. She was one of the first people at Autodesk I was able to ask questions like “what do I wear to work” or “how and when should I reply to emails” being first generation and in the workforce, I appreciated this guidance.
What has been the proudest or most exciting moment in your work so far?
I feel that the proudest and most exciting moment of my work so far has been the moment I felt I owned my title. Because I didn’t study traditional engineering, I did not feel like an engineer.
I was told otherwise by people on my team, and it was very nice and supportive. I was on a mission to prove to myself that I was an engineer.
Recently, we launched a ten-month project where we gave our avatar voice and movement. I was the lead developer for the voice and the conversation interaction with them. During that entire project I felt that I learned a lot — about trying to find creative solutions, managing and communicating issues across different teams, and also engaging with the type of work that I was doing. The moment the project was complete was so exciting, because it was tedious work brought the dialogue around reaching the uncanny valley back.
What advice do you have for young people who are interested in AI-related fields who might just be starting their career journeys?
Include yourself in the dialogue of AI. If you’re interested you should ask questions, even if they’re very simple questions like “what’s the difference between machine learning and AI?” It was difficult for me to start asking those questions. It’s really interesting hearing the different perspectives you get and what you learn. The questions that you have may impact the person that you’ve asked, and they might remember your interest and contact you about it later.
Yizel Vizcarra is a Conversation Engineer on the Machine Assistance (MA) team, focused on improving customer support through dialogue optimization of the Autodesk Virtual Agent (AVA). In her role, Yizel analyzes conversations between Autodesk customers and AVA to inform and enhance AVA’s dialogue design and natural language understanding. In other words, she helps AVA better understand and serve our customers through conversation
Follow along with AI4ALL’s Role Models in AI series on Twitter and Facebook at #rolemodelsinAI. We’ll be publishing a new interview with an AI expert every week. The experts we feature are working in AI in a variety of roles and have taken a variety of paths to get there. They bring to life the importance of including a diversity of voices in the development and use of AI.