An Interview with Scott Prevost, VP of Engineering for Adobe Sensei & Search

“I’ve always been fascinated by how people communicate with machines.”

Some words have never rung more true in today’s technological world. For Scott Prevost, VP of Engineering for Adobe Sensei & Search, he’s been fascinated by how people communicate with machines as long as he can remember. And now, as he leads the Adobe Search team, he gets to implement and improve features that help computers better understand how people communicate their ideas and intentions.

We sat down with Scott to learn about his background and ask a few questions about the team that helps people find exactly what they’re looking for in our Creative, Document and Experience clouds.

Why did you decide to join Adobe?

My fascination with human-computer interaction is what led to me to get a PhD in natural language processing with a focus on building systems that could engage in spoken dialogue. I started off as a researcher, founding a company that made voice-enabled interactive animated characters, and eventually led the team at Powerset, a semantic search startup that was acquired by Microsoft. I spent several years working on search at Microsoft, and then eBay. I’ve had opportunities to run both engineering and product management, which I loved because it allowed me to truly direct the way products get shaped to meet the needs of customers.

This love for product and engineering is what led me to Adobe. I was fascinated by the way Adobe reinvented its business and moved to the Cloud. And the creative product and engineering culture really resonated with me. Since Adobe was just starting its efforts in Search, I knew I would be able to build it from the ground up and champion a data-driven approach to product design with an emphasis on AI and machine learning. This massive potential for transformation excited me.

Can you tell me about the Search team and the technologies your team works on?

My team is tasked with providing search capabilities across Adobe. That means helping customers find all their content—from images, videos, PDFs to content on Adobe.com—by managing search services for products in our Creative, Document and Experience clouds. Our work is centered around deep understanding of digital content enabled by Sensei.

For example, working with Adobe researchers, we’re able to harness 75 million digital media assets on Adobe Stock to build machine learning models. These models understand both the content and concepts represented in images and videos, and their aesthetic characteristics and styles. We work with a variety of data mining and machine learning techniques, from support vector machines to the latest deep learning approaches. We then work with product and design teams to combine these technologies into powerful features and new ways of searching. To enable this, we handle billions of documents and assets, and thousands of events per second, so we’re building scalable, fast search infrastructure to fit our needs.

How does your team affect the customer experience?

Search helps deliver the promise of the cloud. We make our customers’ assets findable whether they’re using desktop products, mobile apps or the web. Since we have a deeper understanding of digital media content than purely storage-focused clouds, we can offer a real advantage for customers keeping assets in our clouds rather than our competitors. Our customers want an easy way to find their own assets as well as assets available in services such as Adobe Stock, and we enable these capabilities right in their normal workflows.

What types of interesting trends in tech are you seeing, and how is your team involved?

There have been so many recent advances in deep learning, which is exactly what our team is riding on. This allows for automatically tagging images, finding images that look the same, generating captions and understanding action in videos. And of course deeper text and spoken language understanding will be really important for search moving forward, especially in the age of Siri and Alexa.

What types of roles are you hiring for your team?

Right now we’re looking for candidates at all levels and experience—from college grads to seasoned managers. We have a number of positions open, including engineers with cloud and services experience, machine learning engineers, data scientists, product managers, program managers, and quality engineers.

What characteristics do you look for?

Because the team is rapidly growing, we want candidates who want to work in a fast paced environment. I think of the team as a startup within Adobe—we’re doing innovative things and there’s a sense of urgency, but we have great resources too.

Ideally we want to recruit people with deep experience with search who are up to date in machine learning knowledge. Candidates with experience in search engineering and cloud based services will fit right in. And of course we look for candidates who are relentlessly customer focused. Perhaps most of all, we value diversity of thought. I firmly believe teams are stronger and more effective when they reflect a variety of viewpoints, backgrounds and skills.

How would you sum up your experiences at Adobe so far?

I’ve been here for almost two years now, and it’s been a great experience. I appreciate the super collaborative environment, and the way Adobe employees really care about the customer. Not to mention, we have amazing products. As the business continues to drive transformation to the Cloud, it’s really an exciting time to be here.

Words of advice for people wanting to join Adobe?

Be prepared to love your job, the people you work with and the customers you serve.

Is Scott’s team a perfect fit for you? Check out our career site and look for Adobe Search positions.