Machine Learning Engineer, Decision Systems (Ad-Tech AI)

An early-stage company is building an autonomous platform that runs digital advertising campaigns, with real budgets at risk. This engineer would own the decision layer that determines what happens to every campaign, while a separate platform engineer handles supporting tools.

Responsibilities

• Develop recommendation and scoring logic for bid adjustments, budget moves, pausing or boosting, and conversion-postback tuning
• Set out the way the platform learns from results so strategy keeps improving rather than restarting
• Write optimization policies for noisy live data, ranging from basic rules to post-training and reinforcement learning
• Plan how several agents reason about a campaign’s situation and cooperate, and assess what they output
• Be answerable for the quality of decisions and the return on ad spend rather than for infrastructure

Qualifications

• Three or more years shipping production ML or optimization systems
• A closed-loop decision system you have delivered, where logged outcomes change later decisions
• Reinforcement learning, bandits or policy optimization applied to a live system
• A quantitative degree such as engineering, physics, math, computer science or statistics
• Statistical reasoning on noisy production data, including experiment design
• Self-directed working style
• Knowledge of ad optimization mechanics (pacing, spend caps, ROAS or CPA targets) is a plus

On-site in the San Francisco Bay Area. Existing visa holders can transfer; new sponsorship is not offered.