We explore topics of data science, machine learning and MLOps. Subscribe below to receive the best bits monthly.
What every data scientist should know about the command lineJuha Kiili
Almost any programming language in the world is more powerful than the command line. Why would you even bother doing anything on it? Don't be fooled: the modern command line is rocking like never before!
Three ways to mitigate model output riskEikku Koponen
Machine learning comes with new types of risk. We need to minimize the risk by addressing how we develop these algorithms and also how we apply these algorithms in the real world. In this article, we'll look at three ways of mitigating the latter – i.e. output risk.
Mike Del Balso joins Valohai’s advisory boardEero Laaksonen
Mike Del Balso is a familiar name to most in the machine learning community. He's one of the pioneers in the MLOps space and has laid the foundations for operational machine learning at Uber, Google, and most recently, Tecton.
Experimentation at Scale: a Q&A with Serg Masís from SyngentaJuha Kiili
Syngenta is a leading provider of agricultural science and technology focused on seed and crop protection products aiming to improve global food security by enabling millions of farmers to make better use of available resources.
One size doesn't fit all - How the use case affects ML system complexityEikku Koponen
Algorithms have become faster, fancier, and more complex in the past couple of years. Still, they haven't gained as much complexity as the systems around algorithms. In this article, we'll discuss three examples of systems complexity.
Docker for Data Science: What every data scientist should know about DockerJuha Kiili
Docker isolates the software from all other things on the same system. A program running inside a "spacesuit" generally has no idea it is wearing one and is unaffected by anything happening outside.
The 3Ps: The foundation of an AI trailblazerEero Laaksonen
People, Processes and Platforms are the foundation for every company looking to be an early-mover in machine learning. Leaders should focus on developing in tandem because unsupported team members will be ineffective and platforms alone can't provide value.
What every data scientist should know about Python dependenciesJuha Kiili
Dependency management is the act of managing all the external pieces that your project relies on. It has the risk profile of a sewage system. When it works, you don't even know it's there, but when it fails, it becomes very painful and almost impossible to ignore.
Valohai strengthens its advisory board with Robocorp CEO Antti KarjalainenEero Laaksonen
We're excited to announce Antti Karjalainen to our advisory board. He's the founder of Robocorp, a leader in developer-first RPA. To Valohai, Antti brings his unique perspective on the developer tooling space and go-to-market strategy.
Managing AI Products: Feasibility, Desirability and ViabilityHenrik Skogström
Product management is as massive a topic as machine learning so let's start with a fundamental question. When is it worthwhile to develop an AI product? A helpful tool most PMs have seen for this is the Sweet Spot for Innovation that IDEO popularized.
Running Weights & Biases Experiments on Valohai PipelinesEikku Koponen
Sometimes it is hard to combine the world of experimenting and the more dev-oriented world of data science with robust pipelines and modular work. This example combines Weights and Biases experiments with Valohai's production pipelines.