Leveraging Argo Workflows and Argo Events for Cloud Agnostic Enter...
summary
A reusable, cloud‑agnostic ML platform is built on Argo Workflows + Argo Events with deployment and bootstrapping via Argo CD and serving on kserve; data landing in buckets triggers thresholded file‑drop sensors that batch files, transform into an offline/online feature store, retrain models, record metrics, and only deploy if metrics improve. Blueprints and Argo templates automate repeatable pipelines, reduce cloud spend, and the kserve inference service example (cars dataset) is validated end‑to‑end with a sample request.
AI-written summary · about
chapters (13)
- 0:00Introduction and agenda
- 1:11ML platform challenges
- 3:01Cloud agnostic approach
- 5:05Argo workflows and CD
- 7:20Application sets on K8s
- 9:48Data preparation strategy
- 11:48Building the feature store
- 13:34Streaming and batch ingestion
- 14:51Operationalizing ML with events
- 17:06Argo event sensors
- 19:23Model training and serving
- 21:27Conclusion and results
- 22:17Q&A session
related talks
26mLeveraging Argo Ecosystem for Scalable Cloud-Native Solutions at Databricks
25mPanel: Argo for ML: Achieving Scalability and User Experience
07mArgo in an AI-Native World: Effective In-Cluster Machine Learning and A...
25mBuilding MLOps POCs and Sandbox Environments Using k3s and Argo
10mScaling Time Series Analysis With Argo Workflows: Patterns and Practi...
38mMachine Learning Pipelines with K3s and Argo – Sergio Mendez
27mEmpowering Autonomy: BYD's Journey Taming Million-Task Scale With Argo...
23mPerception Data Pipeline Reconstruction With Argo Workflows