Frequently Asked Questions

AI in Production: Data Quality & Practical Challenges

What is the main message of the "AI Can't Fix Bad Data" video featuring Zeta Global CTO Christian Monberg?

The video centers on the critical importance of data quality in building AI systems that work reliably in production. Christian Monberg, CTO of Zeta Global, explains that even the most advanced AI models cannot overcome the limitations of poor or inconsistent data. Teams must prioritize data integrity, validation, and governance to ensure AI delivers meaningful results in real-world environments.

Why can't AI fix bad data in production environments?

AI models depend on accurate, consistent, and well-structured data to make reliable predictions and decisions. If the underlying data is flawed—due to errors, missing values, or lack of proper validation—AI systems will produce unreliable or incorrect outputs. The video emphasizes that data quality is a foundational requirement for successful AI deployments, and no amount of model sophistication can compensate for bad data.

What practical steps are recommended for teams deploying AI in production?

The discussion highlights several practical steps: invest in robust data validation processes, implement strong data governance policies, and continuously monitor data quality. Teams should also ensure that their data pipelines are resilient and auditable, so issues can be detected and resolved before they impact AI outcomes. These steps help maximize the reliability and value of AI systems in production.

How does Sedai's approach to safe, autonomous optimization relate to the challenges discussed in the video?

Sedai's platform is designed for safe, autonomous optimization in production environments, emphasizing continuous validation and incremental changes. While the video focuses on data quality, Sedai's technology ensures that optimizations are made without causing incidents or breaching SLOs, complementing the need for reliable data and robust operational practices in AI deployments.

AI Can't Fix Bad Data

In this episode of 1 IDEA, Suresh Mathew sits down with Christian Monberg (CTO @ Zeta Global) to discuss what it actually takes to build AI systems that work in production.

Suresh Mathew

Founder & CEO

June 4, 2026