Principal Product Manager - AI Data Quality

The Mission 

We are building an AI-native enterprise, and high-fidelity data is the substrate. 

We are looking for a technically fluent Product Manager to architect and scale an AI-Ready Data Quality Platform built on Databricks and Unity Catalog. 

This is not a traditional MDM or stewardship role. 

You will define and ship the platform capabilities that make our AI Data Fabric trustworthy, observable, and production-grade — from real-time anomaly detection to CI/CD-native schema enforcement to automated data contract validation. 

If you think of data quality as code, treat governance as infrastructure, and believe AI systems are only as good as the data feeding them — this role is for you. 

What You’ll Own 

Build the AI-Ready Data Quality Platform 

• Define and ship native data quality capabilities inside Databricks Lakehouse 

• Productize policies and controls within Unity Catalog (lineage, access, schema enforcement) 

• Embed data contracts and validation logic directly into pipelines 

• Partner with data engineering to integrate dbt-based transformation layers into quality frameworks 

• Drive metadata, lineage, and semantic standardization as first-class platform features 

Operationalize Data Quality in the AI Data Fabric 

• Design real-time anomaly detection systems (statistical + ML-driven) 

• Build upstream schema validation into CI/CD workflows (shift-left quality) 

• Define SLOs/SLAs for data products 

• Enable automated drift detection for training and inference datasets 

• Implement observability across streaming and batch architectures 

You will treat data quality like SRE treats uptime. 

Drive Data Ownership as a Product Discipline 

• Establish a data product ownership model across service teams 

• Define what “production-grade data” means for AI use cases 

• Build self-service tooling for teams to monitor and certify their data 

• Incentivize measurable quality accountability at the domain level 

This role transforms culture by building the platform that enforces it. 

AI + Governance Convergence 

• Define how governed datasets become AI-ready assets 

• Enable traceability from raw source → curated feature sets → model inputs 

• Align catalog metadata with AI feature stores and inference pipelines 

• Partner with ML teams to support model reproducibility and dataset versioning 

What You Bring 

• 5+ years in Product Management for Data Platforms, Analytics, or AI Infrastructure 

• Deep working knowledge of: 

• Databricks Lakehouse architecture 

• Unity Catalog governance constructs 

• dbt transformation workflows 

• CI/CD patterns for data pipelines 

• Data observability and monitoring patterns 

• Strong SQL fluency and comfort reading Python/Scala data pipeline code 

• Experience defining data contracts and schema evolution strategies 

• Understanding of streaming frameworks (Kafka, Spark Structured Streaming, etc.) 

• Experience supporting AI/ML workloads in production environments 

Bonus: 

• Experience with modern data observability platforms (Monte Carlo, Bigeye, etc.) 

• Familiarity with feature stores and model lifecycle tooling 

• Knowledge of domain-oriented data mesh architectures 

  How We Measure Success 

• % of AI datasets certified as “production-grade” 

• Reduction in downstream model failures due to data issues 

• Automated anomaly detection coverage across critical pipelines 

• Adoption of data ownership model across service domains 

• CI/CD-integrated data validation coverage 

Why This Role Matters 

AI systems amplify whatever data they are fed. 

This role ensures: 

• We trust our data. 

• Our models are reproducible. 

• Governance is automated. 

• Quality is engineered, not inspected. 

You won’t be managing spreadsheets of bad records. 

You will be building the infrastructure that makes AI reliable at scale. 

#LI-JB1