Why can humans intuitively assess data quality, but AI systems still jump to the wrong conclusions?
In this clip from “The Hidden Costs of Poor Data Quality in AI,” a panel hosted by Data Science Connect, Jon Malloy, Senior Technical Account Manager at Snowplow, breaks down why AI can’t fully replicate human context when evaluating data.
Jon explains:
- Why AI can detect anomalies (nulls, bad Unicode, broken rows) but struggles with context and definitions
- How inconsistent definitions of events—like “product views” or “interactions”—lead to poor AI outputs
- Why data governance, shared terminology, and a consistent data model are critical before applying AI
- How misaligned definitions across an organization create hidden costs and unreliable AI decisions
This clip is essential for anyone working in AI, machine learning, data engineering, data governance, analytics, MLOps, data quality, and event data management.
Interested in hearing the full discussion? Catch the webinar here: https://snowplow.io/events/the-hidden-costs-of-poor-data-quality-in-ai
#datagovernance #dataquality #eventdata #datascience AI data quality, data governance, event data, machine learning reliability, data definitions, Snowplow, AI accuracy, data science best practices