iCORE Enterprise MDM / Entity Resolution Platform
iCORE. Many records in. One trusted golden record out.
Databricks-hosted MDM/entity-resolution product combining Next.js workflows, Databricks orchestration, governed review, explainable decisions, and role-based controls.
Databricks BrickbuilderBuilt end to end
- Role
- Data product engineer
- API workflows, Databricks orchestration, governance controls
- Stack
- Databricks + Next.js
- SQL Warehouse, Jobs API, Unity Catalog, Delta control tables
- Scale
- Governed MDM workflow
- Connectors, source mappings, canonical entities, golden records
- Outcome
- Traceable decisions
- RBAC, audit logging, review paths, explainable resolution
The brief. What it took, and what it changed.
Problem
Enterprise source data needs reusable connectors, mapped canonical entities, governed golden records, review workflows, and traceable entity-resolution decisions.
System
Next.js 15 and TypeScript application hosted through Databricks Apps, integrating with Databricks SQL Warehouse, Jobs API, Secrets API, Unity Catalog, and Delta control tables.
Impact
Strongest product-engineering case: owned API workflows, Databricks orchestration, governance controls, audit logging, RBAC, and explainable decisions.
How it works. In 4 stages.
Reusable intake
Connectors and source mappings turn varied source systems into governed onboarding paths.
Master entity model
Canonical attributes, golden records, and history tables shift the product from connector-first to entity-first.
Resolution templates
Match logic, review paths, and explainability make entity decisions traceable instead of opaque.
Databricks control plane
SQL Warehouse, Jobs, Secrets, Unity Catalog, and Delta control tables carry orchestration and governance.
What shaped the build. And what it runs on.
Re-architected from connector-centric workflows into a Master Entity model.
Modeled reusable connectors, source mappings, canonical attributes, golden records, golden-record history, and resolution templates.
Built with
- Databricks Apps
- Next.js 15
- TypeScript
- React
- Zod
- TanStack Query
- Databricks SQL Warehouse
- Jobs API
- Secrets API
- Unity Catalog
- Delta control tables
Proof. What's public, and what's on its way.
References
- Project linkPublic project link to add after approval.Pending
Visual artifacts
- Sanitized UI capturesscreenshotTo approve
- Architecture diagramdiagramTo approve
- Workflow diagramdiagramTo approve
Private project summaryProject detail
Databricks-hosted MDM/entity-resolution product with governed review workflows.
- Source
- Current project inventory and implementation notes
- Confidence
- Medium confidence
- Public note
- Client context, implementation screenshots, and detailed architecture stay anonymized until public approval is confirmed.
Client details anonymizedProject detail
Demonstrates governed orchestration, audit logging, RBAC, stewardship, and explainable entity decisions.
- Source
- Feature inventory mapped to Databricks and product workflow implementation
- Confidence
- Medium confidence
- Public note
- Use public-safe workflow diagrams and high-level feature descriptions until detailed assets are approved.
Tradeoffs. And what stays private.
The tradeoff
Public case study uses high-level architecture and anonymized data because the project is current and client-sensitive.
Kept anonymous
Client and implementation details anonymized until public assets are approved.