Workshop: GenOps: Building a MLOps Platform to Support GenAI Workloads with Open...
summary
Shows an end-to-end GenOps platform deployed on GKE using Kubeflow on GKE, MLflow, Spark on Kubernetes, KServe for model serving, Keycloak auth, Istio ingress and Cert-manager; data pipeline uses embeddings into Chroma (vector DB) and a Streamlit front end with a LangChain RAG flow. Demonstrates a Flan-T5 248M model answering AWS FAQ from vectorized PDFs, proves retrieval prevents hallucination, and explains cost-cutting tactics like GPU time-sharing and model staging on read-only many PVCs; response logging to Label Studio enables DPO feedback.
AI-written summary · about
related talks
31mProductionizing ML with ML Ops and Cloud AI
21mMastering the DevOps Machine Learning Pipeline
41mFrom Prototype to Production Building a Scalable MLOps Platform from a Reference Arch
53mExperience with an end-to-end MLOps pipeline: Practical insights and 3D architectural perspectives
29mBuilding the Bridge Between DevOps and MLOps in a GenAI world
57mDevOps for Machine Learning
42mMLOps 2025: A Journey into the Past and the Future
44mDevOps for Machine Learning - why is it different?