Using Kotlin for Data Science
summary
Claim: one language covers exploration and production for data science and data engineering, demonstrated end to end. Concrete tooling: Kotlin Jupyter kernel, Zeppelin interpreter, JetBrains Big Data Tools, Spark integration, krangl data frames, lets-plot, Kotlin numpy. Surprising numbers and scale: ~7,000 European Slack users, 1 GB of messages processed, 8-node cluster, 42 word clusters from k-means. Built pieces (kernels, plotting, numpy bindings, Zeppelin support) run in live demos, scale to clusters, and show near‑native numeric performance (indexing ~1.5x slower).
AI-written summary · about
chapters (18)
- 0:00Introduction
- 3:45Demo
- 4:32Coding
- 11:38LetsPlot
- 12:42LetsPlot Demo
- 14:55Config JSON
- 16:07Crangle
- 18:47Intersection
- 22:59Similarity Matrix
- 25:34Spark
- 28:23Cluster
- 37:19Big Data Tools
- 38:25Kotlin Libraries
- 40:19Moonpie
- 40:48Performance
- 41:46Python Libraries
- 42:13Summary
- 42:57Roadmap
related talks
41mKotlin for Data Science
50mBuilding Data Science Workflows with Kotlin
47mData Science with Kotlin: Leveraging Deep Analysis with DataFrame
31mData Science Yes, But in Kotlin!
39mKotlin Notebooks: Data Visualization for Developers (No Scientist Required)
51mDataFrame: Kotlin's Innovative Approach to Data Structures
33mBeyond Python: Supercharge your notebooks with Kotlin
47mKotlin – Ready for production