I Thought You Wrote About Data Science…

An Honest Answer to an Honest (and Good) Question

An inquisitive reader sent me a private note asking about my last few blog posts. With permission, I am summarizing the question below:

I’m enjoying your GraphQL posts, but am a little confused. I thought you were focused on data science, so why all the discussion now about an API tool?

First of all, thank you for reading! To your specific question, I have two answers. First, the historical perspective:

In my first blog post, I wrote about what to expect in this blog.

I expect the bulk of what I write to be things I am currently learning how to do. Most of them are likely to be related to Google Cloud, data science, and specifically healthcare data science, but I reserve the right to write about any nerd category of interest.

My *really* long-time readers (from months and months back) will remember early on I had a few different posts that centered around accessing data via APIs — Google Sheets, BigQuery, Census Data, and so forth. So this is by far not my first exploration of the value of APIs in data science projects. And GraphQL is a thing I am currently learning, so here we are.

To the second answer: Data management is a big topic in data science. Access to data, and in particular gaining access to data in ways that reduces or eliminates the need to transform it for use in data science projects, is a really — and I cannot overstate this — **really** big deal. GraphQL warrants more than just a passing overview on my part because of the way it enables significant improvements in the way data is accessed and managed and dramatically simplifies data collection. So as a piece of an end-to-end ML operations solution, GraphQL has the potential to be revolutionary. Thus it very much applies to data science from an overall process and productization standpoint, even if it’s not the same thing as an algorithm to build a predictive model.

Thanks again for the question!