Revolutionizing MLOps: Gorka Erkand on Jozu's Game-Changing Solutions for AI Integration

Beers & Bytes Podcast

What if the key to overcoming AI and ML integration challenges in enterprises lies with one visionary company? Join us as we chat with Gorka Erkand, the CTO of Jozu, who is spearheading efforts to revolutionize the MLOps landscape. Gorka shares his insights on how Jozu's open-source project, KitApps, could be the game-changer in seamlessly packaging AI and ML artifacts. As we enjoy our beers, Gorka opens up about Jozu's strategic use of the Open Container Initiative (OCI) and their innovative Jozu Hub, which together aim to redefine the AI and ML experiences for enterprises, making such integrations a reality rather than a distant goal.
Navigating the complexities of managing large language models (LLMs) and their datasets is no small feat. We explore how Jozu tackles these issues head-on, emphasizing the critical aspects of data versioning, integrity, and security. Discover how custom checksums, data snapshots, and software bills of materials (SBOMs) play a vital role in safeguarding the authenticity and transparency of AI systems. Gorka also highlights the significant advancements Jozu is making in vulnerability scanning and deployment processes, with exciting features like packaging Jupyter Notebooks into microservice containers for easy deployment. Unpack the intricacies of model drift monitoring and the implementation of guardrails, ensuring robust and reliable AI and ML systems that can stand the test of time.

Send us a text

Support the show

Chapter

1. AI and ML DevOps Challenges (00:00:00)

2. Managing Large Model and Data Sets (00:12:12)

3. AI ML Vulnerability Scanning and Deployment (00:21:40)