Quantitative Machine Learning Engineer

We are looking for a strong Quantitative Engineer or Machine Learning Engineer to improve an existing AI system that manages digital advertising campaigns and generates creative concepts. This is not a conventional marketing role. We need someone who can design measurable decision systems, experimentation frameworks, learning loops, and controlled AI agents. ## Technical Responsibilities * Model campaign management as a sequential decision system * Optimize budget allocation, scaling, cutting, geographic targeting, and campaign activation * Connect advertising data with downstream outcome and quality signals * Build experiment and decision logging with complete input snapshots * Define confidence thresholds, evaluation windows, constraints, and rollback conditions * Develop shadow-mode policies before enabling autonomous execution * Separate creative generation from creative performance learning * Represent creatives using structured attributes such as hooks, angles, messaging, visual styles, audiences, and formats * Design exploration and exploitation strategies for testing new creative concepts * Detect performance changes, creative fatigue, and contextual differences * Build natural-language interfaces that compile instructions into typed objectives and approved actions * Integrate advertising platforms, internal APIs, databases, and dashboards * Create auditable and observable agentic workflows ## Required Skills * Strong quantitative reasoning * Machine learning and statistical modeling * Experimental design and hypothesis testing * Python and data-processing pipelines * LLM application development * Claude Code or comparable AI-assisted engineering tools * Agentic AI architecture * API integrations * Production monitoring and observability * Ability to work with noisy, delayed, and incomplete feedback signals ## Strong Advantages * Contextual bandits or reinforcement learning * Bayesian optimization * Causal inference * Time-series modeling * Recommendation or ranking systems * Quantitative trading or automated decision systems * High-volume event-processing systems * Advertising platform APIs * Creative-performance modeling * Feature extraction from text and images ## Application Requirements Please explain: 1. How you would model an autonomous campaign-management system. 2. How you would safely move a decision policy from simulation to production. 3. How you would learn from creatives containing many interacting text and visual variables. 4. How you would balance exploitation of proven concepts with exploration of genuinely new ideas. 5. One relevant system you personally designed or implemented. Generic AI, marketing, or automation proposals will not be considered. We are specifically looking for evidence of quantitative reasoning, experimentation discipline, and production system design.