How to shuffle a big dataset
Addressing the engineering challenge of shuffling a large dataset with a low signal-to-noise ratio.
Deep learning experiments in OCaml
Reproducing classical deep learning results in OCaml using TensorFlow and training models defined in a functional way on a GPU.
Using Kubernetes Pod Metadata to Improve Zipkin Traces
Improving Zipkin traces by utilizing Kubernetes Pod Metadata for debugging latency issues across microservices
OpenAI Scholars 2018: Final projects
Exploring innovative final projects from OpenAI Scholars 2018 cohort.
Hands-Off Deployment with Canary
A blogpost about implementing hands-off deployment using Canary at SoundCloud.
The International 2018: Results
Analyzing the outcomes and performance of The International 2018 esports event.
Large-scale study of curiosity-driven learning
Exploring the impact of curiosity-driven learning on a large scale
OpenAI Five Benchmark: Results
Exploring the latest OpenAI Five benchmark results
What the interns have wrought, 2018 edition
Reflecting on the accomplishments of interns at the end of the season.
Creating Readable Spark Jobs
Improving the readability of Apache Spark jobs for efficient data processing at SoundCloud and other companies.
Learning dexterity
Exploring the intricacies of mastering dexterity in a digital landscape
Variational option discovery algorithms
Exploring variational option discovery algorithms for efficient decision-making in reinforcement learning