IBM MDM ARCHITECT

OTSI•Leawood, KS

2d•Remote

About The Position

Object Technology Solutions, Inc (OTSI) has an immediate opening for an IBM MDM ARCHITECT IBM MDM ARCHITECT (REMOTE ) MAJOR RESPONSIBILITES: · Own data quality, matching logic, and golden record creation · Strong experience in Master Data Management (MDM) · Hands-on with: · Matching & deduplication strategies · Golden record creation · Experience tuning: · False positives / false negatives · Knowledge of: · Phonetic matching (Soundex, etc.) · Data quality rules · Key Responsibilities · Define and tune matching algorithms · Identify and fix: · Duplicate records · Data inconsistencies · Work with business teams to: · Define matching rules · Handle edge cases · Validate new data sources and ingestion logic · Define data quality metrics & SLAs · Good to Have · Experience with tools like: · Informatica MDM / Reltio / IBM MDM · Exposure to SQL QUALIFICATIONS AND EXPERIENCE · The right fit is looking for IBM MDM (virtual) guy who is able to move on-prem MDM to cloud with kubernetes and they would need Java knowledge · Conduct comprehensive reviews of the current MDM ecosystem, including matching and linking algorithm performance, duplicate and suspect processing, and custom extensions. · Evaluate and enhance logging, monitoring, backup, and recovery processes to ensure system reliability and resiliency. · Assess and optimize transaction/query SLAs and index performance, identifying and addressing known issues such as bucket overflows and bulk-load mechanism gaps (including DataStage constraints). · Design, develop, and maintain Java code for MDM algorithms, data processing, and integration components, ensuring code is performant, scalable, and maintainable. · Refactor and optimize existing Java code, and write unit, integration, and system tests for all Java components. · Adjust matching algorithm thresholds to minimize false positives/negatives and add/test new matching attributes (e.g., email + last name, phone combinations). · Assess and implement phonetic matching techniques (e.g., Soundex), and lead feasibility studies to determine the need for single or multi-algorithm approaches for different business lines. · Perform sample pairing for new data sources, collaborating with business teams to identify edge cases and refine algorithm behavior, including store-identifier-based auto-matching where applicable. · Review and enhance frequency-based and dynamic bucketing strategies, assess passive execution options to minimize production impact, and validate interactions with STR tables and existing buckets. · Develop and document bulk load strategies, including ingestion options with and without entity management/event handlers; explore disabling matching for high-volume ingestion scenarios (e.g., 200M+ records) and evaluate alternative ingestion patterns beyond API-only approaches. · Capture and report on metrics, SLAs, and trade-offs for each ingestion strategy. · Target <2 seconds for matching response time through iterative tuning, must-gather data analysis, and leveraging load testing tools (e.g., JMeter); review reconciliation SLAs as part of ongoing optimization. · Address deployment challenges such as outages and unpredictable environment behaviors; explore and implement CICD or improved deployment automation; review and optimize configuration, code, and deployment steps for efficiency. · Plan and implement Auth/PingFed-based authentication, reduce reliance on process IDs and passwords, automate credential rotation, and maintain a dedicated 'Security' category in the scope documentation. · Ensure all code adheres to coding standards and guidelines.

Requirements

Strong experience in Master Data Management (MDM)
Hands-on with matching & deduplication strategies and golden record creation
Experience tuning false positives / false negatives
Knowledge of phonetic matching (Soundex, etc.) and data quality rules
Java knowledge
Java 21 (core and advanced), including object-oriented programming and multithreading
Master Data Management (MDM) platforms and tools
Algorithm development for data matching, linking, and deduplication (including phonetic algorithms like Soundex)
Data integration frameworks and ETL tools (e.g., DataStage)
SQL and database optimization (indexing, query tuning, bulk operations)
Performance tuning and system health monitoring for large-scale data systems
Logging, monitoring, backup, and recovery solutions
Load testing tools (e.g.,JMeter)
CI/CD pipeline tools and deployment automation (e.g., Jenkins, GitHub)
Secure coding practices, including Auth authentication and credential management
Micro services API development and integration (e.g., RESTful services)
Unit, integration, and system testing frameworks (e.g., JUnit, Mockito)
Documentation tools (Confluence, GitHub, DX Guides)
MDM Modernization
Migration from on-premises infrastructure to the cloud
Adoption of containerization for deployment and scalability

Nice To Haves

Experience with tools like Informatica MDM / Reltio / IBM MDM
Exposure to SQL

Responsibilities

Own data quality, matching logic, and golden record creation
Define and tune matching algorithms
Identify and fix duplicate records and data inconsistencies
Work with business teams to define matching rules, handle edge cases, and validate new data sources and ingestion logic
Define data quality metrics & SLAs
Conduct comprehensive reviews of the current MDM ecosystem, including matching and linking algorithm performance, duplicate and suspect processing, and custom extensions.
Evaluate and enhance logging, monitoring, backup, and recovery processes to ensure system reliability and resiliency.
Assess and optimize transaction/query SLAs and index performance, identifying and addressing known issues such as bucket overflows and bulk-load mechanism gaps (including DataStage constraints).
Design, develop, and maintain Java code for MDM algorithms, data processing, and integration components, ensuring code is performant, scalable, and maintainable.
Refactor and optimize existing Java code, and write unit, integration, and system tests for all Java components.
Adjust matching algorithm thresholds to minimize false positives/negatives and add/test new matching attributes (e.g., email + last name, phone combinations).
Assess and implement phonetic matching techniques (e.g., Soundex), and lead feasibility studies to determine the need for single or multi-algorithm approaches for different business lines.
Perform sample pairing for new data sources, collaborating with business teams to identify edge cases and refine algorithm behavior, including store-identifier-based auto-matching where applicable.
Review and enhance frequency-based and dynamic bucketing strategies, assess passive execution options to minimize production impact, and validate interactions with STR tables and existing buckets.
Develop and document bulk load strategies, including ingestion options with and without entity management/event handlers; explore disabling matching for high-volume ingestion scenarios (e.g., 200M+ records) and evaluate alternative ingestion patterns beyond API-only approaches.
Capture and report on metrics, SLAs, and trade-offs for each ingestion strategy.
Target <2 seconds for matching response time through iterative tuning, must-gather data analysis, and leveraging load testing tools (e.g., JMeter); review reconciliation SLAs as part of ongoing optimization.
Address deployment challenges such as outages and unpredictable environment behaviors; explore and implement CICD or improved deployment automation; review and optimize configuration, code, and deployment steps for efficiency.
Plan and implement Auth/PingFed-based authentication, reduce reliance on process IDs and passwords, automate credential rotation, and maintain a dedicated 'Security' category in the scope documentation.
Ensure all code adheres to coding standards and guidelines.