We’re excited to be in Chicago for KubeCon + CloudNativeCon NA 2023! Find us at Booth O20 through the week for live Outerbounds and Metaflow demos, deep dives with our engineers on ML, AI, and data science on Kubernetes, and tons of cool swag! We are also hosting an AI mixer at the Aviary and presenting multiple talks on crafting scalable, user-friendly ML systems at the conference. We are looking forward to seeing old friends and meeting new ones in Chicago!
Outerbounds will make its debut at KubeCon in Chicago this year, and we are excited about the opportunity to engage with you and share some of our learned experiences building ML/AI systems and cloud-prem platforms at scale!
Here’s where you can find us throughout the week:
Monday, November 6
5:05 pm CST
Shri will demo how organizations, big and small, are using Metaflow to rapidly build, test, and deploy GenAI models on top of their Kubernetes-based HPC infrastructure. He will walk you through some of the nebulous gotchas in running planet-scale data-intensive compute workloads efficiently up and down the modern AI stack and share hard-earned wisdom and practical insights for you to take home!
Monday, November 6
9:15 am CST
The practical reality of today’s ML/AI stack is, unfortunately, quite messy. GPU capacity & availability constraints have forced organizations to seek refuge across multiple clouds and invest in on-prem GPU capacity. Cloud data egress economics further complicate this hybrid cloud ML/AI strategy. In this talk, Shri will pull the curtains (for the first time ever!) on how we are thinking about and solving this very real problem in Outerbounds while completely preserving data sovereignty and compute privacy for organizations looking to scale their GPU workloads on sensitive data. This talk will specifically focus on the technical details of cross-cluster, cross-account, and cross-cloud execution of ML workloads and include a spiffy demo!
Wednesday, November 8
5:30 pm CST
Join us at one of the world’s best cocktail bars, Aviary, for a night of multi-course culinary-forward cocktails (by the team at Alinea) as we bring together fellow builders to trade notes, build peer-to-peer relationships, and discuss the latest on AI infrastructure! RSVP here!
Thursday, November 9
11:55 am CST
Social media is abuzz with exciting new AI demos and prototypes every single day! However, moving beyond these prototypes into actual production-ready ML systems that are intertwined with real-world complexity is a rather huge engineering challenge. This talk, by Savin and Yuan, draws on their decades of experience to explore a practical, user-friendly approach to composing & choreographing advanced ML systems that are responsive to real-time changes. If you are curious how ML products like Amazon, Netflix, 23andMe, Zillow work behind the scenes - this might be the talk for you!
We've had a busy summer shipping exciting updates to Outerbounds, all of which help you build ML/AI production systems faster and securely in your cloud account. Thanks to these, you can develop smoothly with cloud workstations while accessing data quickly and securely from data warehouses. You can manage the software supply chain (which matters for ML/AI, too!) and integrate ML/AI into your existing ecosystem via reactive events. All these features work well for traditional ML and data use cases and, of course, for building custom LLM and GenAI solutions, leveraging the lowest-cost GPU resources across clouds. Don’t miss the opportunity to see all these features in action at our booth!
To get in touch with us ahead of KubeCon, reach out here. We are super excited to meet you in person!
Your friends at Outerbounds