Canonical Model & Rule Engineer AI/Agentic Automation

Job Description

Job Description

Remote / WFO / Hybrid: Remote

Hire Type (FTE/Contract): Contract

Project Duration: 5 months – most likely to extend till 12 months

Expected Start Date: 11/02/2026

Rate – 60/hr

Canonical Model & Rule Engineer – AI/Agentic Automation

Experience: 6–9 years
Domain: Healthcare / Medicaid
Project: McKinsey–Health Care Company Medicaid Record-to-Report (R2R) Agentic Automation

Role Summary

We are seeking a hands-on Canonical Model & Rule Engineer to design and implement standardized Medicaid data models, source-to-canonical mappings, configurable business rules, and reconciliation logic for AI-enabled R2R processes. The role will ensure that enrollment, capitation, claims, encounter, remittance, exception, and financial data are consistent, governed, traceable, effective-dated, and reusable across data, backend, agent, and reporting components .

The project model includes entities such as Member, EnrollmentSpan, RateCell, CapitationEvent, RemittanceLine, Provider, Claim/ClaimLine, Encounter/Disposition, ReconciliationResult, Exception, and LineageEvent.1

Key Responsibilities

• Design and maintain the canonical Medicaid data model , including entity definitions, relationships, identifiers, schemas, and data dictionaries.
• Map 834, 820, 837, and 835 transactions and operational source data into standardized canonical entities.
• Define source-to-target mappings, transformation logic, derived attributes, validation rules, and reject-handling requirements.
• Develop configurable and effective-dated business, eligibility, rate-cell, capitation, claims, encounter, and reconciliation rules .
• Maintain reference data, rate tables, mappings, thresholds, state-specific configurations, and rule versions.
• Implement deterministic rules for expected-versus-actual reconciliation, exception identification, materiality, and financial calculations.
• Ensure rule changes are version-controlled, testable, explainable, auditable, and reproducible for historical periods.
• Implement data-quality controls covering completeness, validity, consistency, uniqueness, timeliness, and referential integrity.
• Enable source-to-canonical, source-to-rule, and source-to-metric lineage .
• Build reusable SQL/Python components for transformation, validation, reconciliation, and rule execution.
• Collaborate with Medicaid, finance, data, backend/agent, QA, audit, security, and Controller teams.
• Support unit, data-quality, reconciliation, regression, historical-replay, and rule-validation testing.

Project materials emphasize versioned reference data, member-level reconciliation, claim-to-encounter reconciliation, clause-to-rule traceability, and effective-dated state configuration.2

Required Technical Skills

• Advanced SQL and strong Python
• Canonical, conceptual, logical, physical, and dimensional data modeling
• ETL/ELT, data transformation, normalization, and schema mapping
• Rules-engine and configuration-driven design patterns
• Effective-dated data and slowly changing dimensions
• Reference and master-data management
• Schema evolution, metadata management, and version control
• Data-quality frameworks, validation, exception handling, and reconciliation
• Source-to-target mapping and transformation specifications
• Data lineage, provenance, and metric traceability
• Delta/lakehouse data-engineering concepts
• REST APIs, JSON, XML, XSD, and enterprise data integration
• Automated unit, reconciliation, regression, and data-quality testing
• Git, CI/CD, code reviews, and technical documentation
• PHI/PII protection, RBAC, encryption, masking, and secure data handling

Preferred Technology Experience

• Azure Databricks
• Delta Lake / Delta Tables
• Unity Catalog for metadata, lineage, governance, and access control
• Databricks Workflows
• PySpark and Databricks SQL
• Azure data and security services
• Data-quality tools or frameworks
• Data-modeling tools such as ER/Studio, Erwin, SQLDBM, or equivalent
• Configurable rules engines or decision-table frameworks
• MLflow or equivalent versioning and experiment-tracking capabilities

The project architecture identifies Databricks Workflows, Delta-based data structures, Unity Catalog lineage/access controls, governed functions, and reusable orchestration components as relevant platform capabilities.3 Final Azure, Databricks, and model/platform choices remain subject to Sprint Zero confirmation within the approved enterprise landscape.4

Healthcare / Medicaid Knowledge

Working knowledge of:

• Medicaid enrollment, eligibility spans, and retroactive changes
• Member, provider, plan, product, and state-program data
• Rate cells, member months, capitation, and remittance
• Claims, claim lines, adjudication, payments, and encounters
• State encounter submission and disposition processing
• 834 enrollment, 820 capitation/remittance, 837 claims, and 835 payment/remittance transactions
• Expected-to-actual and claim-to-encounter reconciliation
• State-specific rate books, companion guides, and configuration rules
• Medicaid R2R, exception management, financial close, and control evidence
• HIPAA and secure handling of PHI/PII

The R2R scope specifically connects enrollment and capitation reconciliation with claims and encounter traceability, supported by configurable rules, shared exceptions, and metric-to-source lineage.1

Ideal Candidate

A strong data-modeling and rules-engineering professional who can translate complex Medicaid transactions and finance requirements into standardized canonical entities, configurable deterministic rules, governed reference data, and traceable reconciliation outputs. The candidate should be able to independently design scalable models and rules that are accurate, effective-dated, reusable, testable, explainable, and audit-ready .