inRiver ETL - AI-DLC

Appearance

inRiver ETL PipelineAI-DLC Documentation

Two-Phase Approach: Fast Track + Full Modernization

Requirements Overview
Phase 1 (Fast Track)

Phase 1 - Fast Track (2-3 days)

Python converter with PostgreSQL accumulator. inRiver deltas → Heiler CSVs → Existing Blower

Phase 2 - Full Modernization (4-6 weeks)

Bronze-Silver-Gold-Export architecture with dbt + PostgreSQL, replacing Blower entirely

Strategic Approach

Get inRiver working quickly (Phase 1), then modernize for long-term maintainability (Phase 2)

Project Overview ​

Goal: Migrate from Heiler PIM to inRiver PIM while modernizing Blower infrastructure.

Two-Phase Strategy:

Phase 1: Fast Track (2-3 days) ​

  • Approach: PostgreSQL accumulator + Python converter
  • Flow: inRiver deltas → PostgreSQL → Heiler CSVs → Blower
  • Benefit: Get inRiver working immediately with existing Blower
  • Status: Requirements complete, ready for approval

Phase 2: Full Modernization (4-6 weeks) ​

  • Approach: dbt + PostgreSQL pipeline (Bronze → Silver → Gold → Export)
  • Benefit: Replace Blower with modern, maintainable architecture
  • Status: Requirements documented, deferred until Phase 1 stable

Technology Stack ​

Phase 1:

  • Database: RDS PostgreSQL (accumulator)
  • Processing: Python (~500 lines)
  • Orchestration: AWS MWAA (Airflow)
  • Output: Heiler CSV files

Phase 2:

  • Database: Aurora PostgreSQL (JSONB support)
  • Transformations: dbt (data build tool)
  • Orchestration: AWS MWAA (Airflow)
  • Output: CSV exports

Current Status ​

Phase: Inception - Requirements Analysis
Phase 1: Ready for approval
Phase 2: Documented, deferred
Last Updated: 2026-03-03