About the Project Our client is a leader in AI-powered performance marketing, operating across 25+ verticals with unmatched precision, speed, and scale. Their proprietary technology stack integrates seamlessly with major media platforms, enabling real-time event-level data exchange, optimization, and attribution. At the core of their operation is a deep commitment to AI-driven decision-making. From real-time bidding engines and predictive lead scoring to campaign automation and anomaly detection, their in-house AI models are central to how we scale campaigns, reduce inefficiencies, and outperform market benchmarks. They’ve built and continue to evolve a robust internal platform to empower media buyers, analysts, and operators with real-time alerts, smart recommendations, and semi-autonomous optimization tools.
Role Summary Develops, trains, and evaluates ML models — especially tabular, predictive, and ranking models — and contributes directly to production-first modeling efforts across the funnel.
Required skills: * Strong hands-on experience with tabular ML models (XGBoost, LightGBM, CatBoost, logistic regression, etc.) * Proven experience with data wrangling, feature engineering, dataset preparation * Experience training, tuning, and evaluating models at scale * Solid understanding of model validation * Strong Python skills and familiarity with ML libraries * Ability to translate hypotheses into measurable experiments * Experience working closely with data engineering pipelines
Preferred Qualifications (Nice to Have) * Experience with inference engineering: packaging models, building inference endpoints, optimizing latency * Exposure to model monitoring: drift, data quality checks, performance monitoring * Experience with containerization (Docker) and serving frameworks (FastAPI, Flask, TorchServe, Bento, etc.) * Experience deploying ML models in production environments * Prior experience with ranking, scoring or optimization models
