Who am I?

I have been a professional in the data, machine learning and AI space since graduating with a masters degree in data science from Cardiff University in 2019. In the years since, my exact title has bounced around alot - but with a primary focus on production software rather than research. I have professional experience in the following technologies:

• pyspark / databricks
• fastAPI and REST API design in general
• azure cloud, including a primary focus on kuberenetes
• CICD, model monitoring and automatic refresh strategies for live machine learning APIs
• event based architecture
• agentic and generative AI

What is Cashewe?

The 'cashewe' name plays as an homage to a previous, crudely named `cod cachu` directory in which I built a variety of cheesy prototypes (cod cachu translating to 'shitcode' and often being mispronouned by non-Welsh speakers phonetically as 'cod cashew'). I currently use this github organisation to play with ideas and prototypes that may not be wholey appropriate to invest professional time into, but with a focus on high quality solutions mostly as a matter of pride. In starting this blog my hope is to both share some of the lessons I have learnt from across industries in the space. I expect these to mostly be opinion pieces with occasional looks at how the sausage has been made - im not terribly interested in tutorials however so do not expect these often.

As this is a hobby space:

• AI usage is typically minimal but no garuntees will be made.
• no timelines will be set for new content
• no garuntees of code being bug-free or tested will be made