Accelerating Humanitarian Relief with Kubernetes [I]
summary
Project Fortis ingests Twitter, RSS, Instagram and live radio, runs Spark Streaming on Kubernetes with Kafka and spark native to extract entities, geotags via OpenStreetMap and indexes in Cassandra so dashboards and GraphQL APIs return alerts in <15 seconds. Deployed on Azure ACS/AKS with Helm, uses Cognitive Services for speech and sentiment, Lucene for NER, React frontend, and demonstrates a working multi-cloud Cassandra ring stretched across Azure and AWS.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:30About Erik Christoph
- 3:13The current process
- 3:48Challenges
- 4:33Project Fortis
- 6:12Adaptable Solution
- 7:20Functional Architecture
- 7:59Spark Streaming
- 9:01Cognitive Services
- 9:41Spark on Kubernetes
- 10:10Spark Native
- 11:17Helm
- 11:32Streamlined Developer Experience
- 12:12Deployment Model
- 14:53Azure Container Service
- 16:13Azure Kubernetes Service
- 18:12High Availability
- 21:06Azure Data Centers
- 22:32MultiCloud Cassandra
related talks
01h 47mRadanalytics.io streaming applications workshop
26mOptimal Approaches for Real-Time Machine Learning with Apache Spark on Kubernetes.
46mAdaptive cloud approach: Build and scale apps from cloud to edge
32mKubernetes and AI To Protect Our Forests: A Cloud Native Infrastructure for Wildfire Prevention
28mAccelerating Real-Time Analytics with Cosmos DB and GPU-Enhanced Serverless Apache Spark
26mBuilding Apache Spark Clusters on K8s for AI Workloads
38mHighload — An Unexpected Journey
19mBuild Real time, AI powered Apps