

Data Science "Mix Martial Arts", applied re-inforcement learning, scaling AI workloads using Ray - Max Pumperla - The Data Scientist Show #065
16 snips Jul 28, 2023
AI Snips
Chapters
Transcript
Episode notes
First Project Taught Software Importance
- Max's first startup project built a new ad evaluation system with no existing analytics or software stack.
- He realized to be productive, he needed strong software skills to avoid dependency on data engineers.
Open Source Led to Remote Job
- Max contributed to Keras and created LFS for distributed deep learning with Keras on Spark.
- He landed his first remote Silicon Valley job by tweeting his open source work to Deep Learning 4J creators.
Ray Simplifies Distributed Python
- Ray enables seamless scaling of Python code from laptop to clusters with minimal changes.
- It solves Python's GIL limits by offering simple primitives for parallel and distributed computation.