Blog
- Standalone deployment of Airflow on Kubernetes
- Local development with Kubernetes
- Don't use Airflow, use your CI/CD tool for orchestration
- Using SQLite for choosing a phone to buy
- Dark mode, light mode, balcony
Projects
- Parameter Server implementation on Apache Flink Streaming API. See the code on GitHub.
- Matrix Factorization in Flink: see the code on GitHub for SGD and iALS.
- My Bachelor's thesis: a small machine learning library in Haskell. See the code on GitHub.
Talks
- Journeys from Kafka to Parquet: how (not) to sink a data stream to files.
- At DataWorks Summit 2019, Barcelona. Watch the talk on YouTube.
- At Big Data Spain 2018, Madrid. Watch the talk on YouTube.
- Blogpost. Read it on bol.com's tech blog.
- Parameter Server on Flink, an approach for model-parallel machine learning.
- At Flink Forward 2017, Berlin. Watch it here.
- Building Large-Scale, Adaptive Recommendation Engine with Apache Flink and Spark.
- At DataWorks Summit 2017, München (with Zoltan Zvara). Watch the talk on YouTube.
- Experiments. See the code on GitHub.