How to Mitigate AI Biases Using Data Contracts | Data Governance to Improve Data Quality
We will learn how low quality data leads to wrong predictions, failed language models and AI hallucinations. Data quality and data contracts provide a single surface for collaboration on data in a shared language, allow the data model to evolve in an agile and iterative way, and apply data governance incrementally where it’s needed for AI and ML systems.
Dive into the world of data quality and data contracts — API-based agreements between data producers and consumers that capture the data’s schema, semantics, distribution, and enforcement policies.
About Speaker:
Chad Sanderson is a lead of data platform team @convoy4850 and lead of AI platform team @@Microsoft
We’ll explore how organizations can leverage data contracts to ensure Artificial Intelligence systems are trained on trustworthy and well-governed datasets.
Table of Contents:
0:00 – Introduction
2:00 – What does AI Evangelist imagine?
5:46 – What happens if the data quality is low?
7:38 – How to avoid complex data
12:50 – How to avoid low-quality data
20:46 – Using data contract to avoid AI biases
30:39 – Enforcement mechanism
35:00 – Contract pipeline
35:57 – Monitoring
41:00 – QnA
💼 Learn to build LLM-powered apps in just 40 hours with our Large Language Models bootcamp: https://hubs.la/Q01ZZGL-0
📱 Social media links
Connect with us: / data-science-dojo
Follow us: / datasciencedojo
Keep up with us: / data_science_dojo
Like us: / datasciencedojo
Find us: https://www.threads.net/@data_science...
--
#dataquality #datacontract #AIbiases #datagovernance
