Machine Learning Engineer, Ads

This role is for an engineer who knows how ad systems work, including ranking, targeting, conversion prediction and measurement, and who is also hands-on with generative models. A strong candidate has shipped machine learning systems to production at scale and can move between research-style experiments and reliable engineering. Experience adapting foundation models through fine-tuning, preference methods and careful evaluation matters a lot.

Generative models now create the ad creative itself, so the job shifts from only picking which ad to show to also shaping what gets made, using real performance data as feedback. Much of the routine tuning and iteration can be automated or sped up by models. Humans still decide what counts as a good outcome, design trustworthy experiments, judge creative quality, and own the tradeoffs between advertiser goals and output quality.