Intetics Inc., a global technology company providing custom software application development, distributed professional teams, software product quality assessment, and “all-things-digital” solutions, is seeking a highly skilled and experienced Staff Data Scientist / Machine Learning Engineer to join our dynamic team on a full-time basis.
Responsibilities:
- Design, build, and maintain predictive Lifetime Value (pLTV) models using machine learning techniques such as XGBoost, CatBoost, and probabilistic forecasting methods.
- Develop and manage a tiered scoring framework that assigns users to different prediction models based on data availability and signal quality.
- Create and optimize calibration frameworks that transform raw model outputs into business-ready values through normalization, scaling, and custom loss optimization strategies.
- Own feature engineering processes, building and maintaining scalable feature datasets from transactional, behavioral, and engagement data.
- Define and enforce feature quality standards, ensuring model inputs are reliable, relevant, and suitable for production use. - Monitor model performance, drift, and data distribution changes, proactively identifying when retraining or model adjustments are required.
- Analyze the impact of deployed models on downstream outcomes and detect feedback loops that may affect future model accuracy.
- Develop and improve identity resolution processes to unify behavioral and transactional data across multiple data sources.
- Build, test, and maintain production-grade Python services and data pipelines for model training, scoring, and signal delivery.
- Implement robust software engineering practices, including automated testing, CI/CD, monitoring, idempotent processing, and deployment safeguards.
- Collaborate with data, engineering, and product stakeholders to improve data quality, model effectiveness, and platform scalability.
- Operate and optimize ML workflows in cloud environments using orchestration, scheduling, and distributed processing tools.
- Ensure the reliability, accuracy, and scalability of machine learning systems that support critical business decision-making and performance optimization.
Requirements
- 5+ years of hands-on ML engineering in Python with production-deployed models (not just Jupyter notebooks). You should be comfortable writing production-grade code, not just prototypes.
- Experience with gradient boosted trees (XGBoost, CatBoost, or LightGBM) in a production setting, including hyperparameter tuning, feature importance analysis, and model evaluation beyond default metrics.
- Strong SQL and BigQuery fluency. You will write complex feature engineering queries, scheduled transformations, and data validation logic directly in BigQuery.
- GCP production experience. Cloud Run, Airflow/Composer, Vertex AI, Pub/Sub. You have deployed and maintained ML workloads on GCP, not just completed certification exercises.
- Understanding of model calibration. You know the difference between a model that minimizes RMSE and a model that maximizes business value. You have experience with or strong intuition about how prediction errors compound when consumed by a downstream optimization system.
- Outbound API integration experience. You have pushed model outputs into external systems that act on them (ad platforms, recommendation engines, pricing systems, bidding systems).
Strongly Preferred:
- E-commerce or DTC domain experience. Customer lifetime value, purchase prediction, churn modeling, or customer segmentation for retail or subscription businesses.
- Familiarity with probabilistic customer models (BG/NBD, Gamma-Gamma, or similar frameworks). If you haven't used these specifically, strong Bayesian statistics foundations are equally valuable.
- Ad platform signal engineering. Experience with Google Ads API (ConversionUploadService, Enhanced Conversions), Meta Conversions API (CAPI), or similar server-side event dispatch systems.
- Identity resolution or entity matching. Graph-based identity stitching, probabilistic matching, or CDP-style cross-device identity work.
Nice-to-have:
- Terraform, CI/CD pipeline design, Docker containerization for ML workloads.
- Experience with causal inference methods, A/B testing in ad spend environments, or multi-armed bandit allocation.
- Familiarity with agentic development workflows (Claude Code, Cursor, or similar AI-assisted coding tools).