Keeping Up With Data #55
5 minutes for 5 hours’ worth of reading
Querying data is getting cheaper and faster. We can all discuss if Databricks’ data warehousing performance record is just a marketing stunt, and what are the consequences of a data lake technology leading data warehousing benchmark. But seeing data storage and querying prices dropping is a good news for an uninvolved observer.
What’s not a very good news is a data education gap at the top of organisations. Research by Profusion is shoving that CEOs tend to have worse understanding of data concepts than directors and managers. In times, when every company is a data company, it would be great to see the CEOs leading the ranking. The research validates my anecdotal experience why data initiatives are often driven by other C-suite members.
So let’s work on our understanding of data concepts with the following articles.
- Data and the almighty dollar: Data industry is booming. There are more and more categories and vendors forming the modern data stack. And as these companies individually are persuading customers how commercially beneficial they all are, the data bill is growing. While its tempting to see it as a sign of a bullish trend in data industry, and accept it as costs of doing business, it’s not to only option. And I see it every day, that companies embarking on a data journey are not enthusiastic to onboard 20+ technologies all with aggressive ARR targets reflected in their pricing models. I agree with Benn that the market will likely consolidate driven by the customers’ effort to consolidate their data expenses and pragmatically decide on what is actually needed to do the job they need to do. And what will leave them with the biggest profit margin. (Benn Stancil)
- Data Mesh — Fad or Fab? Data mesh seems to be one of the dividing emotional topics. People either love it, or hate it. At least those who are vocal about the topic. See for instance the discussion on LinkedIn when this article was published. I’m a big proponent of data democratisation, data fluency, and using data to maximise its economic value. Does data mesh help with that? I still don’t know. Maybe it’s too early to say. I’m still left confused and puzzled after reading anything about data mesh. Or maybe I’m limited by my vocabulary, which doesn’t allow me to process statements like: “There are autonomous data product quantums that provide multimodal access to domain data for analytical workloads — connected together in a graph — each both transforming and serving/controlling immutable bitempiral data.” So, data mesh — fad or fab? I’d rather skip the question for now. (Barr Moses @ TDS)
- The lost Art of Data Modeling: With the changes in the data infrastructure and general approach to data by many companies, the role of data engineers has been changing rapidly. Traditionally, data engineers were primarily focused on data modelling — ‘shaping the data by developing an understanding of the underlying data and the business process’. In time, most of the work shifted towards managing data pipelines, from ingesting data into a data lake, designing and operating DataOps and MLOps processes to dealing with real-time processing of data streams. The data engineers became ‘data movers’. But even though the data modelling skill is no longer the primary skill, it is still very relevant. (Julien Kervizic @ Hacking Analytics)
I’m talking at a conference in Prague next week about data as an asset and its value. Kind wait to give a talk in Czech (it’s been some time), see old colleagues and meet new ones!
