Calculating the Value of Pie: Real-Time Survey Analysis With Apache Kafka®
summary
Builds a Telegram bot that writes survey entries (Avro, schema registry) into Kafka, uses ksqlDB to create streams and a histogram aggregation, and outputs both real‑time and a 48‑hour tumbling window final result. Uses compact topics for respondents, emits changes vs emit final to control updates, and pushes the final verdict back to the phone via a Kafka Connect HTTP sink. Pipeline runs on Confluent Cloud, produces live counts and a final shopping list — it works (pumpkin and apple pies win).
AI-written summary · about
chapters (32)
- 0:00Introduction
- 1:20What are you baking
- 4:41The problem statement
- 5:50Surveys are everywhere
- 7:59Survey analysis in general
- 9:12Making it more exciting
- 10:22What is Apache Kafka
- 13:19Event Platform
- 14:40Pie Example
- 18:12Kafka Architecture
- 23:09Kafka Cluster Example
- 27:07Where to Start
- 28:27Understanding the Data
- 29:24Kafka Schema Registry
- 33:02Kafka Survey IDs
- 34:24Write Survey to Kafka
- 36:32Conversation Handler
- 37:14Message Handler
- 38:14Send
- 40:55Processing Data with Kafka
- 44:46Define Inputs
- 46:19Normalize Data
- 48:53Survey Responses
- 49:12Create Table
- 50:17Create Histogram
- 51:05Output
- 51:24RealTime Results
- 51:53Late Results
- 52:56Kafka Connect
- 55:18Summary
- 56:26Baking something
- 58:46Stream processing use cases
related talks
59m🤖 Building a Telegram bot with Apache Kafka and ksqlDB
41mApache Kafka and ksqlDB in Action: Let’s Build a Streaming Data Pipeline!
29mApache Kafka and ksqlDB in Action : Let’s Build a Streaming Data Pipeline
31mFun with Sensors
57mTaming IoT Data: Making Sense of Sensors with SQL Streaming
46mBuilding Streaming Data Apps with Pac-Man
58mRealtime Analytics Geofencing Kafka
49mLet’s Build a Streaming Data Pipeline!
