I've given a number of talks at conferences and other meetups over the years. Here are some of the materials for a few of them. My presentations tend to be less slide-dependent than some, so these materials are not guaranteed to be useful. If you're certain I've given a talk which you don't see here, but want access to the materials, let me know -- I may have them, I may not.
I've given several talks (some multiple times) on machine learning:
Earlier this century, bayesian classifiers were all the rage. I gave a talk at a few BarCamps explaining Bayes' Theorem (an old S5 presentation, so functionality not guaranteed).
Another popular talk was "Write your own Bayesian Classifier!", which I'm ashamed to say I sometimes tried to do in a 30-minute slot.
I've given a brief survey on machine learning concepts and techniques.
I once gave a lightning talk about SVMs.
I've given several talks on PostgreSQL. Most talks I've collected here have videos of the presentations, rather than slides. The quality of the videos varies, sorry.
A few years back I gave a talk on using postgres_fdw to federate queries.
I did a bit of a deep dive into postgres partitions.
A while back, I gave a talk about the current state of using postgres in cloud environments. Cloud environments have changed significantly in the intervening time, so I can no longer stand by my recommendations in this talk.
For the local PUG, I gave a somewhat tongue-in-cheek talk about two locking problems I had encountered.
I've given a number of other talks, some for local python user groups, some about other dynamic language issues. For a while, the Portland Python User Group had a "module of the month" feature. Here are two that I presented:
A comparison of profile and cProfile
A brief overview of the multiprocessing library