Senior Data Scientist and AI Specialist

Deloitte Deloitte · Consulting

Senior Data Scientist and AI Specialist role focused on designing, building, and deploying production-grade AI/ML systems, including NLP, RAG, forecasting, and advanced analytics. The role involves fine-tuning and operationalizing LLMs and RAG pipelines, establishing MLOps on Azure, and developing data models for AI/ML integration, particularly within regulated health & human services environments.

What you'd actually do

  1. Design, build, and deploy production-grade AI/ML systems across NLP, Retrieval-Augmented Generation (RAG), forecasting, and advanced analytics use cases.
  2. Fine-tune and operationalize large language models (LLMs) and RAG pipelines, including embedding strategy, vector store design, retrieval evaluation, and prompt/grounding patterns that meet the accuracy, traceability, and data-handling standards required in regulated health & human services environments.
  3. Establish and maintain end-to-end MLOps on Azure (e.g., Azure ML, pipelines, model registry, CI/CD, monitoring, and drift detection), ensuring models are reproducible, secure, observable, and reliably maintained in production rather than one-off prototypes.
  4. Develop SQL data models and analytics solutions that integrate and structure data from child welfare and Medicaid program systems, translating fragmented agency data into trustworthy inputs for AI/ML systems and operational reporting that improve efficiency and surface actionable insight.
  5. Partner with cross-functional teams, product owners, agency stakeholders, engineers, and policy/program staff — to translate public sector business problems into technical solutions, and to clearly communicate model behavior, limitations, and outcomes to both technical and non-technical audiences.

Skills

Required

  • Python
  • SQL
  • Microsoft Azure
  • NLP
  • RAG
  • LLMs
  • MLOps
  • data modeling

Nice to have

  • forecasting
  • advanced analytics
  • embedding strategy
  • vector store design
  • retrieval evaluation
  • prompt/grounding patterns
  • CI/CD
  • monitoring
  • drift detection
  • child welfare data
  • Medicaid program data

What the JD emphasized

  • production-grade AI/ML systems
  • regulated health & human services environments
  • end-to-end MLOps on Azure
  • accuracy, traceability, and data-handling standards

Other signals

  • production-grade AI/ML systems
  • fine-tune and operationalize LLMs and RAG pipelines
  • end-to-end MLOps on Azure
  • integrate and structure data for AI/ML systems