Data Scientist - Modelling

JPMorgan Chase JPMorgan Chase · Banking · Metro Manila, National Capital Region, Philippines · Consumer & Community Banking

Build and improve machine learning models and analytics that translate data into measurable business outcomes. This role focuses on planning, developing, and executing analytical projects, partnering with developers for implementation, and evaluating model performance. Experience with generative AI, agent-based workflows, and RAG systems is required.

What you'd actually do

  1. Plan, develop, and execute analytical projects using appropriate techniques
  2. Partner with developers to implement models and analytics deliverables and integrate enhancements into existing solutions
  3. Evaluate and monitor performance of new and existing models using defined metrics and thresholds
  4. Explore new data sources, techniques, and environments to enhance model accuracy and robustness
  5. Communicate model results and analytical findings through clear presentations and visualization

Skills

Required

  • Master’s degree in statistics, computer science, engineering, or a related quantitative field or bachelor’s degree plus three years of relevant experience
  • Experience with generative AI, including building agent-based workflows and retrieval-augmented generation systems
  • Working knowledge of XGBoost, random forest, and k-means algorithms
  • Advanced proficiency in data analysis and machine learning techniques
  • Intermediate programming proficiency in SQL and Python or R
  • Intermediate experience with data visualization tools (for example, Matplotlib or Tableau)
  • Experience applying supervised and unsupervised learning methods in practical use cases

What the JD emphasized

  • Experience with generative AI, including building agent-based workflows and retrieval-augmented generation systems

Other signals

  • building agent-based workflows
  • retrieval-augmented generation systems
  • XGBoost, random forest, and k-means algorithms
  • supervised and unsupervised learning methods