Luke Arrigoni started Arricor in 2012 to help large companies make sense of their data. Since then, he and the team have taught organizations like Goldman Sachs, AT&T, and Thomson Reuters about the principles of AI. His secret? Focus on the business problem and the right technology approach becomes obvious.

Listen and learn...

1. How UPS uses AI to automatically assign the right tax code for packages
2. What responsibility AI developers have for the decisions their algorithms make
3. How to clean dirty data to make it ready for AI model training 
4. When to use neural nets vs. gradient-boosted trees
5. Which tasks are good candidates for classifier models vs. NLP
6. Which job skills are future-proof... and which are likely to be replaced by automation 

References in this episode:

• Fish from Mozart Data on AI and the Future of Work
• Airflow for data pipeline automation