Bridging the Gap: Real-Time Predictive Analytics with Faustream
summary
Claims a working open‑source pipeline called Faustream that wires Python models into Kafka + Faust for low‑latency inference, not theory. Configuration uses a plug-and-play YML config to point at input/output topics and model paths; demo shows live sentiment JSONs flowing through Faust processing to a pre-trained model and back to Kafka. Tool is lightweight, framework-agnostic about model formats, currently tuned for text streams, and plans include dynamic model reloading and wider connector support.
AI-written summary · about
chapters (13)
- 0:00Welcome
- 0:47Presenter and Faustream
- 1:10Real-time analytics challenges
- 2:37Faustream features
- 4:02Kafka and Faustream architecture
- 5:49Setting up and configuring Faustream
- 7:13Live sentiment analysis demo
- 8:17Real-time analytics use cases
- 9:23Faustream advantages
- 10:49Challenges and future plans
- 12:38Open source and Q&A
- 13:27Compute costs and edge devices
- 16:03Closing
related talks
51mSimplifying Event-Driven Services with FastStream: A How-To Guide
47mBuilding High Performance Real-time Stream Processing Microservices
14mManaging decisions with Amazon Kinesis for better outcomes
46mComposable microservices for streaming analytics
41mMastering real-time anomaly detection with open source tools
57mMachine Learning Data Pipelines
30mEvent Streaming and the Future of Applications
36mStream All The Things!