Senior Machine Learning Engineer – Customer Analytics & Recommendation Systems

Ubique Systems; Helsinki; Presencial; EUR 85 000 - 120 000; Tempo integral

Resumo da oferta

Ubique Systems in Helsinki is seeking an experienced ML Engineer to develop, train, validate, and deploy models for churn, propensity, LTV, and recommendations. You will collaborate with data engineering and business teams to build scalable feature stores and predictive pipelines using modern MLOps.

The role emphasizes end-to-end model development, feature engineering, time-based validation, and deployment within a lakehouse architecture on AWS (SageMaker, MLflow, Kubeflow).

Qualificações

• Experience building ML models for churn, propensity, and lifetime value.
• Experience with feature stores and predictive pipelines.
• Collaborating with data engineers and business teams to translate requirements into models.
• Proficiency with MLOps tools and cloud platforms.

Responsabilidades

• Develop machine learning models for churn, upsell/cross-sell propensity, LTV, and next-best-action recommendations.
• Translate business definitions into model targets, eligibility populations, and prediction horizons.
• Define target/label logic for churn, revenue contraction, acquisition, renewal, and customer value use cases.
• Collaborate with data engineers to build training and scoring datasets from governed feature stores.
• Perform exploratory analysis, feature selection, leakage assessment, and time-based validation.
• Develop sentiment-based features from customer text and assess their impact on models.
• Generate and rank next-best-action recommendations using model scores and business priorities.
• Validate model performance across time periods and customer segments.

Conhecimentos

ML/AI concepts

Feature engineering

Model validation

Time-based validation

Sentiment analysis

Ferramentas

SQL

PySpark

Python

AWS S3

Kubeflow

MLflow

SageMaker

Dremio

Iceberg/Delta

Descrição da oferta de emprego

ROLE OVERVIEW

We are seeking an experienced ML Engineerin Helsinki.

This role focuses on developing, training, validating, and implementing machine learning models for customer churn, propensity, lifetime value, and recommendation engines. The ideal candidate will work closely with data engineering and business teams to construct scalable feature stores, build predictive pipelines, and leverage modern MLOps frameworks.

2. TECHNICAL COMPETENCIES NEEDED

• SQL
• Pyspark
• ML-DL-NLP
• AWS S3
• Python
• PySpark
• Machine learning model development
• training and validation
• Feature engineering and target or label creation
• AWS S3 and data-lake or lakehouse architecture
• Apache Iceberg-Delta table and Dremio experience
• Kubeflow
• MLflow
• SageMaker
• or equivalent MLOps experience
• NLP modeling

3. KEY RESPONSIBILITIES

• Develop machine learning models for customer churn, upsell propensity, cross-sell propensity, downsell or right-sizing, customer lifetime value, and next-best-action recommendations.
• Work with business analysts, product owners, and Data Engineers to convert business definitions into model targets, eligibility populations, prediction horizons, and measurable customer outcomes.
• Define target and label logic for customer churn, product churn, revenue contraction, product acquisition, product upgrade, renewal, and customer-value use cases.
• Work with Data Engineers to build point-in-time-correct training and batch-scoring datasets from governed feature-store definitions.
• Perform exploratory analysis, feature selection, leakage assessment, class-imbalance handling, model development, probability calibration, and time-based model validation.
• Develop and validate sentiment-based features from eligible customer free text and assess their contribution to churn and propensity models.
• Develop next-best-action candidate generation and recommendation-ranking logic using model scores, customer value, expected margin, product eligibility, customer benefit, and business priorities.
• Validate model performance and stability across industry, company size, customer-value bands, product portfolios, and time periods.