Super Data Science: ML & AI Podcast with Jon Krohn 879: Serverless, Parallel, and AI-Assisted: The Future of Data Science is Here, with Zerve’s Dr. Greg Michaelson
63 snips
Apr 15, 2025 In this session, Greg Michaelson, Co-founder of Zerve and former Chief Customer Officer at DataRobot, sheds light on innovative features like the fleet parallelization for efficient coding. He discusses how AI tools, especially large language models, might threaten existing SaaS companies. Greg's unique ‘good-cop, bad-cop’ approach enhances LLM responses, and he dives into the potential of retrieval-augmented generation (RAG) for powerful AI applications. He provides insights into the future of developer workflows and the rising need for flexible AI solutions.
AI Snips
Chapters
Books
Transcript
Episode notes
Inaccurate Recommendations
- Greg Michaelson shared how a media recommendation system using LLMs wasn't accurate.
- It recommended different versions of Dune when asked for similar movies, highlighting the need for better context.
Simplified Deployment
- Zerve's built-in API builder and GPU manager simplify model deployment.
- Data scientists can deploy models directly, removing the dependency on software engineers.
Zerve Free Tier
- Get started with Zerve's free tier, which offers flexibility with resource caps.
- The free tier was temporarily disabled due to Bitcoin miners, but it's now back with new controls.
