* Hiring on behalf of our client
Title: Senior ML / Optimization Engineer
Time: Full-Time
Location: San Francisco, CA (On-site - 5 days/week)
Compensation: $200,000 - $300,000/year + competitive equity
The Company
Our client is an AI-native startup building autonomous technology for performance marketing. Its platform uses AI agents to optimize paid acquisition across major advertising channels, making real-time decisions around bidding, budgets, campaign performance, and spend allocation. The company has grown to approximately $30M in revenue with 30% month-over-month growth and is backed by top venture capital investors. The team is small, highly technical, and operates with significant autonomy, with founders and technical leadership remaining directly involved in building the product.
Role Overview
Our client is seeking a Senior ML / Optimization Engineer to own the intelligence layer behind its autonomous growth platform. This is not a traditional ML infrastructure role. You will build the systems that determine what happens to live advertising campaigns, developing recommendation, scoring, and optimization policies that learn from outcomes and improve future decisions. You will work on production systems managing real advertising spend, with direct accountability for decision quality and performance.
Key Responsibilities
• Build recommendation and scoring systems that drive decisions around bid changes, budget reallocation, campaign pausing, boosting, and postback optimization.
• Design closed-loop learning systems where logged outcomes feed back into and improve future decisions.
• Develop optimization policies over noisy, real-world production data, ranging from rule-based strategies to reinforcement learning and post-training approaches.
• Apply reinforcement learning, bandits, and/or policy optimization to live production systems.
• Design and orchestrate multiple AI agents that reason over campaign context, make decisions, and evaluate outcomes.
• Develop experimentation frameworks and apply rigorous statistical reasoning to measure decision quality.
• Continuously improve how the system learns from previous outcomes rather than resetting strategy between campaigns.
• Take ownership of optimization performance and the real-world returns generated by the decisions your systems make.
• Work autonomously alongside a highly technical, hands-on founding and engineering team.
Qualifications
• 3+ years of experience shipping production machine learning or optimization systems.
• Quantitative STEM degree in Computer Science, Engineering, Mathematics, Statistics, Physics, or a related quantitative discipline.
• Demonstrated experience building and shipping a closed-loop decision system where observed outcomes feed back into future system decisions.
• Hands-on production experience with reinforcement learning, multi-armed bandits, and/or policy optimization applied to a live system.
• Strong statistical reasoning, including experience designing experiments and working with noisy production data.
• Strong Python experience and production ML engineering capabilities.
• Ability to independently identify problems, make technical decisions, and ship without requiring detailed direction.
• Comfortable working in a fast-moving environment where models and optimization strategies have measurable real-world outcomes.
Preferred Experience
• Experience with ad optimization mechanics such as bid pacing, budget allocation under spend caps, ROAS or CPA targets, and/or postback optimization.
• Experience designing or evaluating agentic systems and multi-agent workflows.
• Experience with recommendation or scoring systems where decisions continuously adapt based on observed outcomes.
• Experience working in an early-stage, high-autonomy startup environment.
What This Role Is Not
This role is unlikely to be a fit for candidates whose experience is primarily:
• Advertising ML or ranking systems without meaningful agent, reinforcement learning, bandit, or policy-optimization experience.
• LLM applications limited primarily to RAG, prompting, or chatbot development without production decision or optimization systems.
Why Join
• Own the intelligence, not the plumbing: A separate platform engineer owns simulator and platform tooling, allowing you to focus on optimization policy, learning loops, and decision quality.
• Work against real outcomes: Your systems will make decisions involving live advertising spend, with performance measured directly against returns.
• Direct access to technical leadership: Work within a flat, approximately 20-person team alongside hands-on founders and technical leadership.
• Applied optimization at scale: Work across reinforcement learning, bandits, post-training, agent orchestration, and rule-based strategies in a commercial production environment.
• High autonomy: The culture emphasizes ownership and results rather than micromanagement.
Compensation & Location
• Base salary of $200,000 - $300,000/year.
• Competitive equity package.
• Full-time, on-site in San Francisco 5 days per week.
• Visa transfers (including H-1B and OPT transfers) are supported; new visa sponsorship is not available.