Welcome to our blog series, “UW Researcher Spotlight.” These posts highlight University of Washington researchers who, as principal investigator—or “PI”—lead their teams on the path from idea to impact with the support of CoMotion. The resulting innovations have been licensed to industry, spun off as a startup, or licensed direct to the user from within the UW.
Stay tuned for profiles about UW researchers from computer science & engineering, life sciences, emerging tech, social sciences and more.
Amrita Mazumdar: CEO, Vignette AI
Luis Ceze, Advisor; Mark Oskin, Advisor
When did you file your first Record of Innovation (ROI) with CoMotion?
Tell us about your innovation – what technologies have you licensed or spun off?
We spun off the core technology behind the Vignette research project into a company, Vignette AI.
What problem does your innovation solve?
Vignette AI improves user experience on low-bandwidth video connections and reduces costs for video distribution, by combining machine learning with traditional video streaming software services. Our technology uses machine learning to predict which areas of a video are most visually important and eliminates degradation of perceived video quality under bandwidth constraints, and the resulting videos can be played with any kind of codec or video player.
What impact are you having?
We are in the early stages of rolling out our technology in a pilot program (please connect if you want to try out Vignette AI!). Vignette AI reduces costs for video distribution services without impacts to cloud workflow or user experience, and we’re excited to bring these cost and experience benefits to market.
What funding have you received?
We received a CoMotion Innovation Gap Fund in Fall 2019, and we recently completed the Jones + Foster Accelerator program through the Buerk Center for Entrepreneurship. I’m supported right now through the CoMotion Commercialization Fellows program. The Gap Fund was critical in walking us through our first customer discovery efforts and developing our software prototypes. As part of the CoMotion Fellows Program, I’ve been able to focus on developing those early customer relationships, building our company’s business operations and early funding, and learning from the other fellows in the program.
What’s the latest?
We’re excited to start our pilot program in the coming months, and well as some of the partnerships we are setting up so far. We are looking to work with companies dealing with high costs from video distribution or struggles with quality-of-experience delivering video to low-bandwidth areas, so if those problems resonate with you please reach out!
Do you have any commercialization advice?
Customer discovery is the most important aspect of commercialization. Even if you don’t know what kind of impact your technology might have, talking to people in different industries and learning about their pain points will only serve to make your technology and its impact stronger. Even once you have a better understanding of the customer problem you’re solving, keep talking to your customer! I have learned that a close relationship with your target customer is the biggest motivator for keeping the flywheel of your business moving.
What other CoMotion services and resources have you used and how did they help?
I attended the Idea-to-Plan workshop, Customer Discovery workshops through the Innovation Gap Fund, and pitched at DubPitch. As I said before, I’m also part of the Commercialization Fellows Program.
What do you like most about working with CoMotion?
As a solo founder, I have really benefited from the generosity of support and availability from the whole CoMotion team. Every time I have felt “stuck” growing into the next stage of commercialization, CoMotion has provided advice, resources, and connections to helpful people in the community.
Tell us a fun anecdote about your journey on the path to commercialization.
One of the best parts of my “job” is getting to watch video for “work”, but one TV show I don’t watch is Silicon Valley (it hits too close!). From the many, many people who bring it up though, I hear that it is about a startup working on compression in some way? Like I said, it hits very close to home, and you would not believe how many people reference the show when I tell them about Vignette AI.
Who is your favorite scientist or researcher in your field?
Lynn Conway has always been an inspiration to me, not just in her technical achievements that shaped the way we design computers today, but also in her enthusiasm and tenacity at all the different roles she has held in her career.
Can you share with us a bit more about yourself?
I am a computer science researcher by training, with interests in new systems for virtual reality (VR), video, and graphics. I received my PhD in computer science from UW in 2020. During graduate school, I did research in the computer architecture group, the UW Reality Lab, and the UW databases group, and I also collaborated with Facebook Reality Labs and Google Research. Through my industry collaborations, I learned about exciting innovations happening around VR in video processing and machine learning for visual perception, and I saw that this work wasn’t really being incorporated in computer systems or hardware architectures. This influenced my graduate research to focus on building systems that could support a future of perception-optimized video, one result of which is Vignette.
Lab website: www.vignette.company
Follow us on Twitter: @vignette_ai
Subscribe to our newsletter to stay up to date about programs, resources and events.
About UW CoMotion: CoMotion at the UW partners with the UW community on their innovation journey, providing tools, connections, and acumen to transform ideas into economic and societal impact.