The Answer to Life, the Universe, and Everything
summary
Claims that only ~15% of big-data projects reach production (Gartner) versus many teams hitting near 100% readiness by building repeatable platforms. Shows thousands of nodes, three people running clusters with near-zero downtime, hundreds of apps on one multi-tenant platform, and a data fabric spanning 50+ data centers. Argues Kubernetes is the de facto app manager but punts state, so teams need a “Kubernetes for data”; describes a decade-old production data platform and productized model deployment, management, and monitoring that actually ships and runs.
AI-written summary · about
chapters (5)
related talks
34mThe Data Analytics Platform or How to Make Data Science in a Box Possible
35mCloud Native Data Platforms - The why, what and how
09mKeynote: Kubernetes for GenAI: The New Operating System of the AI-First World
09mPlatform Engineering for Data Science at Scale
35mEvolving and Supporting Stateful, Multi-Tenant Decisioning Applications in Production
19mBuilding Production-Ready RAG Systems
42mThe Challenges Managing a Kubernetes-Based Machine Learning Infrastructure
32mDesign Patterns for Running AIML and Bigdata Workloads on Kubernetes