ML Platform Engineer

About the role

The ML Platform Engineer will build MCP Servers for CIE, integrate with AgentCore Gateway, deploy BHI custom models to Databricks Model Serving + Snowflake, implement AWS Bedrock integration for LLM explainability, implement PHI guardrails, manage MLflow model registry, build CI/CD for model deployment, and integrate Feature store.

Responsibilities

• Integrate with AgentCore Gateway (JWT/mTLS authentication, entitlement checks, routing)
• Deploy BHI custom models to Databricks Model Serving + Snowflake (UDFs, external functions)
• AWS Bedrock integration for LLM explainability via AI Marketplace subscription
• PHI guardrails implementation (pre-request scanning, post-response masking for healthcare data)
• MLflow model registry (versioning, A/B testing, rollback)
• Build CI/CD for model deployment
• Feature store integration (Databricks Feature Store MCP tools)
• OpenTelemetry instrumentation (usage events emission to AI Marketplace metering)
• MCP tool manifest registration (score_claim, explain_detection, provider_profile tools)

Requirements

• AWS/ Databricks Certified (preferred)
• Python 5+ years
• Familiarity with MCP Protocol (Model Context Protocol)
• Experience with MLOps/MLflow (3+ years, model registry, serving)
• on AWS and Databricks
• Databricks Model Serving production (REST API endpoints)
• Snowflake UDF + external function deployment
• OAuth2/JWT authentication + mTLS certificate handling, AWS Bedrock (LLM integration for explainability)
• AgentCore Gateway integration (JWT validation, entitlement enforcement, tool routing)
• PHI/HIPAA compliance (pre/post-request PHI scanning)
• OpenTelemetry (usage event emission, distributed tracing)
• REST API development (MCP tool endpoints)
• Docker/containerization
• Monitoring (Databricks, CloudWatch)
• Git + GitHub + CI/CD
• 3-5 years MLOps with AWS + Databricks + Snowflake
• Understanding of MCP Protocol or willingness to learn (AWS AgentCore documentation)
• Worked with OAuth2-secured APIs

Qualifications

• Must have 3-5 years MLOps with AWS + Databricks + Snowflake
• Must understand MCP Protocol or willingness to learn (AWS AgentCore documentation)
• Must have worked with OAuth2-secured APIs

Skills

• AWS
• Databricks
• JSON Web Token
• Rest API Development
• CICD
• Docker Containerization

Benefits

• Paid vacation
• Paid sick leave
• 10 paid holidays
• Bereavement leave
• Jury duty
• 401(k) Retirement Plan with employer matching
• Medical, dental, and vision insurance
• Accidental death and disability insurance
• Short- and long-term disability benefits
• Health Savings Account (HSA)
• Flexible Spending Account (FSA)