Principal AI Engineer

Verizon Verizon · Telecom · Basking Ridge, NJ +2

Principal AI Engineer at Verizon focused on building, training, and deploying complex propensity models to predict customer behaviors, churn risk, and lifecycle triggers. The role involves end-to-end model development, collaboration with marketing stakeholders, and technical mentorship, aiming to optimize customer journeys and reduce churn.

What you'd actually do

  1. Building, training, and deploying our most complex and high-priority propensity models to predict customer behaviors, churn risk, and key lifecycle triggers.
  2. Operating independently to manage end-to-end model development pipelines, from advanced feature engineering to production deployment and monitoring.
  3. Collaborating closely with high-visibility base management pods and senior marketing stakeholders to translate complex business retention goals into actionable data science solutions.
  4. Owning the lifecycle of models deployed in high-impact areas, ensuring continuous optimization, accuracy, and measurable business performance.
  5. Translating advanced analytical outputs into actionable microsegmentation strategies that enable marketing partners to deliver highly personalized customer experiences.

Skills

Required

  • Python
  • R
  • SQL
  • advanced statistical modeling
  • large-scale data extraction
  • machine learning
  • propensity models
  • production deployment
  • model monitoring

Nice to have

  • Master's degree in Data Science, Computer Science, Statistics, or a highly quantitative field
  • LLM agent architecture
  • Google ADK
  • LangChain/LangGraph
  • multi-agent orchestration
  • retrieval-augmented generation (RAG)
  • MCP
  • A2A protocols
  • ONNX-based embeddings
  • vector databases
  • semantic search
  • cloud-scale data engineering
  • BigQuery pipelines
  • GCP (Cloud Run, Vertex AI)
  • customer retention strategies
  • churn prediction
  • customer lifetime value (CLV) modeling
  • A/B tests
  • production ML/LLM monitoring
  • observability tooling
  • OpenTelemetry
  • Arize Phoenix
  • Galileo
  • model drift
  • accuracy degradation
  • output quality
  • agile, pod-based operating model
  • cloud platforms (GCP, AWS, or Azure)
  • production MLOps tools
  • communication skills
  • presentation skills

What the JD emphasized

  • Four or more years of experience independently developing, deploying, and optimizing complex machine learning or propensity models in a production environment.
  • Experience working on high-priority or high-visibility technical initiatives within a corporate setting.

Other signals

  • predict customer behaviors
  • churn risk
  • customer lifecycle
  • microsegmentation
  • hyper-personalization
  • customer retention