alibi-detect VS Zulip

Numerous tools exist for detecting anomalies in time series data, but Alibi Detect stood out to me, particularly for its capabilities and its compatibility with both TensorFlow and PyTorch backends.

That is, to see whether a test image belongs in the distribution of the training images and to provide a routine for special cases. After a bit of reading Ive found that this is related to the field of drift detection in which I tried out alibi-detect . Whereby the training images are trained by an autoencoder and any subsequent drift will be flagged by the AE.

Wasserstein distance is not very suitable for drift detection on most problems given that the sample complexity (and estimation error) scales with O(n^(-1/d)) with n the number of instances (100k-10m in your case) and d the feature dimension (192 in your case). More interesting will be to use for instance a detector based on the maximum mean discrepancy (MMD) with estimation error of O(n^(-1/2)). Notice the absence of the feature dimension here. You can find scalable implementations in Alibi Detect (disclosure: I am a contributor): MMD docs, image example. We just added the KeOps backend for the MMD detector to scale and speed up the drift detector further, so if you install from master, you can leverage this backend and easily scale the detector to 1mn instances on e.g. 1 RTX2080Ti GPU. Check this example for more info.

Seldon | Multiple positions | London/Cambridge UK | Onsite/Remote | Full time | seldon.io

At Seldon we are building industry leading solutions for deploying, monitoring, and explaining machine learning models. We are an open-core company with several successful open source projects like:

* https://github.com/SeldonIO/seldon-core

* https://github.com/SeldonIO/mlserver

* https://github.com/SeldonIO/alibi

* https://github.com/SeldonIO/alibi-detect

* https://github.com/SeldonIO/tempo

We are hiring for a range of positions, including software engineers(go, k8s), ml engineers (python, go), frontend engineers (js), UX designer, and product managers. All open positions can be found at https://www.seldon.io/careers/

I work in this area and also contribute to outlier/drift detection library https://github.com/SeldonIO/alibi-detect. To tackle this type of problem, I would strongly encourage following a more principled, fundamentally (statistically) sound approach. So for instance measuring metrics such as the KL-divergence (or many other f-divergences) will not be that informative since it has a lot of undesirable properties for the problem at hand (in order to be informative requires already overlapping distributions P and Q, it is asymmetric, not a real distance metric, will not scale well with data dimensionality etc). So you should probably look at Integral Probability Metrics (IPMs) such as the Maximum Mean Discrepancy (MMD) instead which have much nicer behaviour to monitor drift. I highly recommend the Interpretable Comparison of Distributions and Models NeurIPS workshop talks for more in-depth background.

All of the above functionality and more can be easily used under a simple API in https://github.com/SeldonIO/alibi-detect.

Zulip — Real-time chat with a unique email-like threading model. The free plan includes 10,000 messages of search history and File storage up to 5 GB. also, it provides a self-hostable open-source version.

(1) Zulip Chat - https://zulip.com/ - seems to be reasonably popular, but more people should know about it

I’ve been using it for over 5 years now [1], and it’s as good as ever. It’s way faster than any other chat app I’ve used. It has a good UI and conversation model. It has a simple and functional API that lets me curl threads and write blog posts based on them.

(only problem is that I Ctrl-+ in my browser to make the font bigger – I think it’s too dense for most people)

(2) re2c regex to state machine compiler - https://re2c.org

A gem from the 90’s, which people have done a great job maintaining and improving (getting Go and Rust target support in the last few years). I started using it in 2016, and used for a new program a few months ago. I came to the conclusion that it should have been built into C, because C has shitty string processing – and Ken Thompson both invented C AND brought regular languages to computing !!

In comparison, treesitter lexers are very low level, fiddly, and error prone. I recently saw dozens of ad hoc fixes to the tree-sitter-bash lexer, which is unsurprising if you look at the structure of the code (manually crawling through backslashes and braces in C).

https://github.com/tree-sitter/tree-sitter-bash/blob/master/...

These fixes are definitely appreciated, but I think it indicates a problem with the model itself.

(based on https://lobste.rs/s/endspx/software_you_are_thankful_for#c_y...)

[1] https://www.oilshell.org/blog/2018/04/26.html

A few more truly in the vibe of open source projects not advertising their hosting providers: https://plane.so/ , https://element.io/ , https://www.loomio.com/ , https://zulip.com/ , and it keeps going... Very few open source projects, in the FOSS sense, are advertising their hosting provider.

I was so excited to see this happen!

I'm not a customer of yours, but your blog posts inspired me a lot. Your journey through quitting caffeine is a great and heartening read.

I've got two things to say;

1) Will you consider source-availabling the web portal (app.keygen.sh) too? Some enterprises could use it for easy management/support for custoner's licenses. Although now that I think about it, it could also discourage custom, more suitable implementations for each use-case... I'm torn on this one. I would like to see it available on GitHub too just out of curiosity too. It's very beautiful.

2) For a team + customers' chat, I cannot recommend Zulip enough. It's a joy to use and has the most innovative chat system I've ever seen. https://zulip.com

I hope your business keeps prospering!

Zulip | Senior Flutter Engineer | REMOTE or San Francisco | Full-time | https://zulip.com/

At Zulip, we’re out to build the world’s best collaboration platform, and we’re committed to keeping it 100% open source. Zulip is the only modern team chat app that is designed for both live and asynchronous conversations. Our product serves as the communication hub for businesses, open-source projects, educators and communities around the world.

We're building the next generation of Zulip's mobile apps in Flutter. We're looking for a senior engineer with Flutter experience to join our small core team and help define the future of team chat. Our Flutter prototype is just a few months old, so this is a greenfield opportunity to help shape the app's architecture from early on.

For full details, check out https://zulip.com/jobs/. Apply at [email protected].

Anyways, I'm an internet stranger, not a social media expert. So let me know what you all think. And if we make a discord or zulip or something to make this a reality, let me know and I'd love to help any way I can.