The company is a small engineering team building a production recommendation system that picks ads in real time.
You would be the first engineer, working across data infrastructure, machine learning and serving systems, with a direct effect on advertiser results and the user experience.
Responsibilities
Build low-latency retrieval, ranking and reranking systems that choose campaigns and creatives for each ad opportunity, and ship them.
Create data pipelines and feature infrastructure to train models on click, install and conversion signals, which arrive on different timelines.
Build user and context representations from engagement history, conversation data and session signals to personalize ads.
Design serving infrastructure that is cheap to run and meets sub-second latency targets.
Build generative agents that produce playable and video ad variants, and feed their performance back into the recommendation system.
Requirements
You have 0 to 3 years of experience and show real technical depth across backend infrastructure, machine learning and AI agents.
You have shipped an ML or AI system to production and owned it end to end, from data pipelines and features through models to serving.
You reason about systems from first principles and can design data architecture, pipelines and infrastructure, not just models.
Experience with recommendation systems, ranking, AdTech, PyTorch or consumer-scale infrastructure is valuable.
Active GitHub work or side projects showing backend, ML or infrastructure skills are a plus, as are game development or familiarity with mobile games and reward ads.
You must be a US citizen or green card holder and able to work on-site five days a week.
Conditions
The base salary is $200,000 to $250,000 per year, plus equity.
Visa sponsorship is not available.
The role is on-site in San Francisco, California, 5 days a week.
Benefits
Equity is part of the compensation package.
Where you’d work
From the office, 4 or more days a week in the office