Product Manager, Data Ingestion & Quality

Protege is a company focused on solving the challenges of accessing the right training data for AI. They are seeking a Product Manager to enhance their data ingestion platform, ensuring data quality and consistency across various verticals. The role involves defining product requirements, overseeing validation processes, and collaborating with cross-functional teams to improve data usability.

Responsibilities:

• Define the stages, validation gates, and quality checks that data passes through from partner arrival to catalog-ready; own the platform requirements that make this repeatable across modalities and verticals
• Own the product decisions around what metadata gets extracted or generated at ingestion, including transcripts, tags, confidence scores, schema inference, at what threshold, and how it gets stored and surfaced
• Define what "catalog-ready" means, build the tooling that enforces it, and get into the data directly to validate that standards are being met; you’ll run queries and review pipeline outputs, not just read dashboards
• Work with vertical stakeholders to translate their "what does ready mean for our vertical" requirements into consistent platform-level standards that don’t require custom engineering per deal

Requirements:

• 4–7 years of PM experience where the core product was a data pipeline, data quality system, or data ingestion platform — you've owned the 'raw data in, trusted data out' problem before
• Hands-on technical depth — you can write SQL, read pipeline logs, spot a schema mismatch, and understand the tradeoffs in a data validation architecture; you look at data directly to verify things are working, not just at metrics
• Experience with external data — you've worked on a product that ingested messy, inconsistently formatted data from third-party partners and had to make it trustworthy; you know what that problem actually feels like
• Build-versus-partner judgment — you've made vendor decisions in a fast-moving technical domain; you know how to evaluate a tool against requirements that will change, and how to structure relationships that preserve flexibility
• Cross-functional credibility — you'll be writing requirements that multiple engineering teams and vertical PMs depend on; you can hold a technical conversation and a product conversation in the same meeting
• Experience with data quality frameworks, metadata standards, or catalog tooling, including dbt, Great Expectations, data contracts, or similar
• Familiarity with de-identification approaches for sensitive data — PHI, PII, or confidential enterprise data
• Background in healthcare data operations, financial data infrastructure, or any domain where data quality has real downstream consequences
• Exposure to ML training pipelines or AI data workflows, where data fitness affects model outcomes
• Experience with data governance strategies