Jonathan Sedar

Personal website and new home of The Sampler blog

Don't call it a comeback

Posted at — 18 Oct 2019

Writing is a great way to learn, and also to connect with people. So when I started out on my own in 2013 and built Applied AI1, I made sure to maintain a regular technical blog - called The Sampler2.

I drafted in my colleagues, and over the course of about 3 years we published a good number of market observations and in-depth technical posts spanning: Bayesian inference with pymc3; semi-parametric survival analysis; feature reduction and high dimensional data viz; options volatility pricing;
forensic blockchain analysis; and data science team management strategies.

During 2017, a combination of major new client project, team restructuring, and the ICO boom3 took all my available brainspace, and the blog went fallow.

I'm pretty sure by now blogging must be incredibly old-fashioned, but I still search out and read a good number of technical blogs, and I miss the mental exercise of building reproducible research to learn and share ideas.

So I've (re)started this blog under my own name to reconnect with folks, rehome some of my old posts from The Sampler, and to keep writing new material as I hopefully continue to learn and apply interesting statistical techniques to the day job.

I'd better get on with it…


  1. Applied AI Ltd was a niche data science consultancy launched back in the days when that was still a fresh idea. Myself and two colleagues provided bespoke data analysis, statistical modelling and expert advice to a range of financial services companies in UK and Ireland.

  2. A moderately nerdy reference to MCMC sampling that I still shamefully enjoy

  3. I got quite into blockchain analysis and graph theory - and wrote up some of that work at acropora.io