PyCon Balkan 2018 Talks
We need millions of images, and we need them yesterday
In this talk we analyze how AWS S3 + Lambda helped us grab image metadata (such as resolution, dominant color and optimal cropping) from tens of millions of images, from various datasets, saving our client a good chunk of $ and us the pain of dealing with the usual "boring" concurrency issues (workload distribution, synchronization, etc.).
All done in, of course, good ol' Python ;)