search

ThomasFaria / retex-innovation-insee

https://thomasfaria.github.io/retex-innovation-insee/authorsample.pdf
0 stars 4 forks source link

Part 5 : new profils (ML engineers etc) will also emerge due to the training evolutions #33

Open mpjoubertdebellefon opened 2 months ago

mpjoubertdebellefon commented 2 months ago

In the same spirit as my previous issue in the introduction, a point about the sentence "recruiting profiles that are not typically present in NSOs, such as data engineers or ML engineers, could bring new essential skills that lie at the intersection of statistical methodology and computer technique" Actually these profiles are not present at the moment in NSOs because of the age pyramid and the training people received. Since these new skills are now (more) taught in the schools where NSOs recruit their agents, there is not a need to radically change the way NSOs recruit their agents, this evolution will be (certainly, not perfectly) brought by the renewal of agents. Maybe the sentence could be slightly modified to incorporate this idea, such as "the emerging trend towards changing work profiles in NSOs, with the increasing share of data engineers and ML engineers (also linked to a generalized evolution in young recruits training), might bring new essential skills that lie at the intersection of statistical methodology and computer technique"

avouacr commented 1 month ago

In the same spirit as my previous issue in the introduct

And for the same reason I'm not sure to agree, I think it is a partial view to let this process being determined only by the age pyramide or the initial training people receive. There can be conscious choices to modify the profiles being recruted, as illustrated for instance by the INSEE-DINUM report stating the needs for much more data scientists / engineers in the public sector

RLesur commented 1 month ago

I think we've to be careful not to generalize too much about what we know at INSEE. INSEE has made a choice, which to my knowledge is unique in the world, to recruit only one profile for all its functions.

In this context, we are highly dependent on the evolution of initial training and the age pyramid.

Other NSIs also have an age pyramid problem, but when it comes to recruitment, they don't tie their hands like we do. They don't hesitate to diversify the profiles they recruit at university. We could decide at INSEE to recruit some data engineers or developers, for instance, but we don't. We do prefer keeping the same channels for recruiting.