1003- Founding ML Engineer

We need an English-fluent Founding ML Engineer, based in Latin America, available to work remotely.
We are looking for someone with exceptional communication and relationship-building skills, who embraces changes while maintaining strong attention to detail. An interested and proactive person, who's constantly learning and improving their skills.

Are you the one we are looking for?

Responsibilities:

• Design, build, and deploy end-to-end LTV prediction systems, covering:
  Data ingestion, Feature engineering, Model training and evaluation and Deployment and monitoring
• Develop ML approaches that work within adtech constraints, such as:
  Delayed or sparse conversions, Noisy attribution, Changing platform policies
• Own MLOps, including: Reproducible training pipelines, CI/CD for models, Logging, monitoring, and alerting. 
• Data quality checks and drift detection.
• Collaborate closely with Product and GTM teams to translate business goals (profitability, payback, repeat rate) into model objectives.
• Help define and evolve the company’s Signal Engineering playbook: What signals are computed, How often they’re updated and How they’re delivered to downstream systems.

Requirements:

• Strong foundations in machine learning, with the ability to reason from: Business objective, data limitations, model choice and deployment
• Hands-on experience building production ML systems (not just notebooks), including: Training pipelines, Deployment and Monitoring.
• Experience with LTV modeling, such as: Probabilistic or BTYD-style approaches, Survival or retention modeling, Regression/classification for value prediction and Model calibration.
• Comfortable working with modern data stacks and cloud environments.
• Core: Python, SQL, Docker (2–4 years experience).
• Data & Warehouse: BigQuery, dbt-style transformations, event and transactional pipelines (Shopify, CRM, GA4, CDPs)
• Cloud: GCP (Cloud Run, Pub/Sub / queues, scheduled jobs), secure APIs and services.
• Machine Learning: scikit-learn, XGBoost / LightGBM / CatBoost, optional PyTorch And MLflow or Weights & Biases for tracking.
• Orchestration: Airflow, Prefect, Dagster (approach matters more than the tool)