Full Job Description
Job Description
What You Will Own
Medallion Architecture - Bronze through Gold
• Design, build, and operate the full medallion stack on Databricks: raw landing (Bronze), identity-resolved and feature-engineered Silver, and the governed serving layer (Gold) that the Helm agent layer queries on every conversational turn
• Implement Delta Live Tables (DLT) and Lakeflow pipelines for batch and structured-streaming ingestion of transaction signals, behavioral events, and third-party identity data
• Own schema enforcement and evolution at the Bronze boundary so upstream change surfaces as a controlled event, not a downstream failure
• Build the unstructured processing tier (Bronze Silver) using Databricks AI Functions or ai_query to convert PDFs, brand guidelines, compliance rule sets, and creative assets into governed Silver tables
Unity Catalog & Governance
• Own Unity Catalog as the governance layer: access control, lineage tracking, and per-client tenant isolation across Helm's financial institution client base
• Implement PII governance controls at the pipeline layer - redacted ID egress, consent signal propagation, and guardrail alignment to GLBA, Fair Lending, and UDAAP
• Build and maintain data-quality expectations frameworks: reconciliation, deduplication, and lineage checks that reduce manual investigation cycles
Serving Layer & Platform Operations
• Build and operate the Gold serving layer behind Helm's twelve analytical models: MTA, MMM, incrementality, LTV:CAC, reach/frequency, cohorts, cobrand overlap, channel viability, card activation, product holdings, scenario reallocation, and channel universe
• Meet the serving contract the agentic layer depends on: fixed signal shapes, live query on every turn, predictable latency, and per-client isolation
• Own CI/CD, infrastructure-as-code, cost management, and pipeline observability for the Databricks environment
Required Qualifications
• 5-7 years of data engineering experience with 2+ years of production Databricks work: Delta Lake, Unity Catalog, DLT/Lakeflow, structured streaming, and SQL warehouses
• Strong Python for transformation logic, data-quality automation, and pipeline orchestration
• Expert-level SQL; comfort designing schemas for both analytical and serving workloads
• GCP fluency: BigQuery, Cloud Storage, Cloud Run, IAM - Helm runs on GCP and Databricks runs within it
• Hands-on experience with at least one clean room or privacy-preserving data exchange technology
• PII governance knowledge in a regulated industry context: data residency, consent frameworks, GLBA, Fair Lending, UDAAP
Preferred Qualifications
• Databricks AI Functions (ai_query) or Vertex AI used inside pipelines for document extraction and normalization
• dbt or a comparable transformation and lineage framework
• Identity resolution concepts at the pipeline layer: deterministic matching, RampID or UID2 linkage, match-rate monitoring
• LiveRamp, AMC, GMP, or XMi clean room connector experience
• Familiarity with LangSmith or LangGraph as a data consumer - understanding what the agentic layer needs from the serving layer
The base compensation range for this role in the posted location is 86129 - 127189
Capgemini offers a comprehensive, non-negotiable benefits package to all regular, full-time employees. In the U.S. and Canada, available benefits are determined by local policy and eligibility and may include:
• Paid time off based on employee grade (A-F), defined by policy: Vacation: 12-25 days, depending on grade, Company paid holidays, Personal Days, Sick Leave
• Medical, dental, and vision coverage (or provincial healthcare coordination in Canada)
• Retirement savings plans (e.g., 401(k) in the U.S., RRSP in Canada)
• Life and disability insurance
• Employee assistance programs
• Other benefits as provided by local policy and eligibility