Vice President, Research, Yahoo DSP
-
IT/Tech
Job Description & How to Apply Below
Yahoo DSP is a rapidly growing demand-side platform, and it runs one of the largest real-time decision systems in advertising. Every time someone opens a news story, starts a streaming show, plays a podcast, or walks past a digital billboard, an auction runs, and it’s over in about a tenth of a second. We bid in those auctions hundreds of billions of times a day across connected TV, video, display, audio, native, and digital out-of-home, for advertisers who spend billions of dollars a year with us.
Each bid is a science prediction, an economics-defined price, and a control-system derived pacing decision made under uncertainty, at a scale that few science teams ever touch.
We have signals most of the market doesn’t: purchases from Yahoo Mail (more than $25 billion in receipts a month), intent from Search, and interests from Yahoo’s content, tied to 213 million authenticated users through Yahoo ConnectID. DSP Research turns those signals into outcomes. We use machine learning to value impressions when conversions are sparse and delayed, bid just enough to win, pace budgets against uncertain supply, and find the cleanest path to every publisher.
We need to build a research organization that will rebuild the decision engine for a world that is increasingly connected and connected through connected TV.
You are a scientist who has built and run a research organization that ships. You’ve published, you still read the papers, but you are deeply embedded in AI. You’re a teacher, a business owner, and you love to dive into the numbers. You question every requirement and conclusion, especially your own. You are a passionate collaborator with engineers and believe anyone can do real science with enough support.
You’ll lead a significant global DSP Research organization and grow it. You’ll set the research strategy with Product and Engineering, decide what we will and won’t study, and defend the time it takes to do work that pays off in quarters, not sprints. As the most senior researcher in the DSP organization, you’ll help set business priorities and lead research initiatives across the broader Yahoo science community.
Everyone in the DSP Engineering team is an agentic builder, and you’ll help us reinvent how research gets done: you’ll lead your team to experiment with agents that propose hypotheses, run cheap tests, and create synthetic data. You’ll build a transparent organization where people feel safe to challenge the same requirements and conclusions that you challenge.
- Own the DSP research strategy and roadmap. Choose investments by business impact, size of the capability gap, and where science creates differentiated value. Say no to the rest, out loud and with reasons.
- Turn research into production. Partner with engineering so new models, controllers, and bidding strategies move from offline results to A/B tests to 100% of traffic, with clear validation standards and rollback.
- Set the scientific bar for DSP: experimentation and evaluation methodology, offline-to-online validation, and the statistical standards everyone uses to decide whether something worked.
- Build the organization. Hire scientists and research leaders, grow the leaders you have, define research career paths and evaluation so researchers aren’t graded on sprint velocity, and set our publication and external‑engagement policy on purpose.
- Protect exploration. Fund a named share of research capacity for longer‑horizon work, such as identity, audiences, and connected TV prediction, and keep it from being consumed by the urgent.
- Lead agentic research: the harness, the skills, and the metrics that prove research throughput and quality are going up.
- Lead and improve frontier research and papers…
(If this job is in fact in your jurisdiction, then you may be using a Proxy or VPN to access this site, and to progress further, you should change your connectivity to another mobile device or PC).
View / Apply for Jobs
Matching My Jurisdiction