MDM Solution Architect

Job title: MDM Solution Architect – Warehouse-Native Identity Resolution (Snowflake / SQL)
remote – India
2 WIDE open
Contract: C2H
Experince:7+ Years
See below and make sure recruiting team understands the request
This is not an MDM platform configuration role. The client is building a custom MDM hub on their own warehouse and is not using Informatica, Reltio, Profisee, or similar platforms. Candidates whose primary experience is configuring vendor MDM products have consistently been rejected.
The ideal candidate must be able to design an MDM solution from first principles and should be comfortable starting with a blank sheet of paper.
This includes:
Defining entities, relationships, and party models
Designing identity resolution strategies
Building match, merge, and survivorship logic
Designing reversible merge processes
Creating golden record methodology
Writing and reasoning through the SQL that implements the solution
Working in a warehouse-native environment using SQL, dbt, and custom data models
A major differentiator is the ability to explain how the solution was built, not just what it does. Many candidates have been able to discuss MDM concepts, tooling, and architecture at a high level, but struggled when asked to explain implementation details, data model design decisions, matching algorithms, survivorship rules, or how they personally built the underlying logic.
The client has repeatedly emphasized that they are looking for candidates who have built rather than configured.
Additionally, most candidates evaluated so far have had experience supporting organization or customer master data. The client's most critical need is experience with person-level identity resolution at consumer scale, involving millions of records where:
Reliable identifiers may not exist
Matching relies on probabilistic techniques
Over-merges create privacy and compliance risks
Identity resolution accuracy is business-critical
Candidates should be able to speak deeply about:
Entity and relationship modeling
Identity resolution
Match scoring and confidence scoring
Survivorship strategies
Golden record creation
Merge/unmerge logic
Data lineage
Warehouse-native implementations
SQL-based solutions
What has not worked well:
Candidates who primarily administer or configure MDM platforms
Candidates who stay at the conceptual or governance level
Enterprise architects who cannot explain implementation details
Data architects whose experience is primarily strategy, oversight, or program leadership
Candidates who discuss what a tool does instead of how the underlying logic works
What has worked best so far: The strongest candidate interviewed was able to discuss:
Building a Client Master across 22+ source systems
Custom match and merge logic
Identity resolution concepts
Confidence scoring
Attribute-level survivorship
Golden record generation
Reversible merge design
Data lineage and source traceability
Even then, the client is seeking deeper evidence of recent hands-on implementation experience, which remains one of the most important criteria for this role.

Work arrangement
Yes