smedegaard.io

I've been reading up on MLOps. I attended talks and enrolled to Coursera Machine Learning Engineering for Production (MLOps) Specialization. But the best source so far has been the book "Builing Machine Learning Pipelines". It's a treasure of practical and applicable methods to build different kinds of ML pipelines. But

I recently took the Deep Learning Institute course "Fundamentals of Accelerated Data Science with RAPIDS" RAPIDS is a framework that let's you move your data handling and machine learning to the GPU instead of the CPU. More iterations, more innovation!

* Space and rockets are cool! * AI is cool! * Data engineering is cool! So what could be cooler than combining the them!? Companies like Glint Solar, DHI GRAS, and Lumen Energy use satellite images combined with artificial intelligence to make a positive impact on the world. Using  Convolutional Neural Networks (CNNs)

I recently expanded and documented my knowledge in the field of Computer Vision. Completing the Udacity Computer Vision Nanodegree, I used deep learning techniques—from basic image processing, to building and  customizing convolutional neural networks. In the exam projects I applied these concepts to vision tasks such as automatic image

I recently got acquainted with Apache Beam which is a batch and stream processing tool. It can be run on different runners but Google Cloud Dataflow is probably the most widely used. I went on to familiarize myself with the Google Cloud data engineering environment and lastly got certified as

I'm a firm believer that we need to do something about the climate crisis. But after following obvious advise like reducing heating in the house and use of fossil fuels, it's not always easy to know what to do or how. Tibber and Tomorrow Luckily there are people working to

After about 6 months of working with Apache Kafka I Thought it was time to get a certificate to confirm that my knowledge on the subject had reached a professional level. Besides the knowledge I've accumulated by working with Kafka I used Linux Academy's preparation course to study for the

"In the previous episode..." I wrote a short post on MQTT and VerneMQ some months ago and promised a follow-up. Here it is. Scope of the post In this post I'll explain how to set up a VerneMQ server on a Ubuntu virtual machine on DigitalOcean. We will use LetsEncrypt

I've been writing Elixir for a while now, but I haven't published a public Hex packages, until now.

I've been playing around with MQTT for a while now and despite the extensive documentation of the Tortoise library I made some (noob 😶) mistakes that had me scratching my head for quite a while. The scope of this article I'll not describe what MQTT is. There's several good resources online

In recently finished RabbitMQ In Depth by Gavin M. Roy. Apache Kafka has been quite hyped for a while now, but RabbitMQ can solve some of the same problems and comes with OTP-goodness since it's written in Erlang. Log vs. queue Kafka uses an immutable log whereas RabbitMQ is a

November 7th 2018 I attended a tutorial named "Thinking concurrently". It was a part of the Code Mesh conference in arranged by Code Sync. It was very interesting to hear Armstrong's insights and anecdotes. Thanks for the opportunity to ask and learn from one of the inventors of Erlang and

Christmas is over, and so is January. Almost anyway. I've finally gotten around to finish Fred Hebert's book on Property-based Testing (PBT) and it's great! By reading this book you get an introduction to PBT, some theory and applied examples with both Erlang and Elixir code. So if you are

My first attempt at embedded Elixir