Conjecture? Palantir is known to be rabidly expensive. The available information on the internet and from people I know who've encountered the technology directly describe their "product" as a mix of a very outdated tech stack and expensive integration services with expensive on-site support [1]
My guess is that the NYPD did an analysis of the life-cycle cost of the system over another contract period, and probably determined they could do something cheaper some other way. Since they've built another system based on a competitor's technology (with some special home rolled glue sauce) they're obviously getting some benefit from the basic approach, but given a choice of close enough tech and more badges or the status-quo they've decided to move forward.
1 - I've basically heard their software described as a big graph backend with a document store and an ancient Java client for mucking around with the graph. Most of their tech has been subsumed by very free and highly scalable technologies.
The most recent video shows a just normal run-of-the-mill web app any decent team of 3 or 4 good web devs could crank out in a few weeks.
They actually don't have any video up of anything recent that actually really shows their core product. The video here shows a java client https://www.youtube.com/watch?v=5OYy_UtINo4 but it's from 2015.
Kind of a shame considering how much money and talent they've vacuumed out of the market. If I'm right, it's amazing for IBM to actually come in at a lower price than some competitor.
I also am not aware, but given this is the NYPD, isn't there an opportunity for the developers to become unionized?
And I wonder, the software developers for the larger trade unions, do they get under the union umbrella or do they get the short shrift form the unions they would work for?
My guess is that the NYPD did an analysis of the life-cycle cost of the system over another contract period, and probably determined they could do something cheaper some other way. Since they've built another system based on a competitor's technology (with some special home rolled glue sauce) they're obviously getting some benefit from the basic approach, but given a choice of close enough tech and more badges or the status-quo they've decided to move forward.
1 - I've basically heard their software described as a big graph backend with a document store and an ancient Java client for mucking around with the graph. Most of their tech has been subsumed by very free and highly scalable technologies.
The most recent video shows a just normal run-of-the-mill web app any decent team of 3 or 4 good web devs could crank out in a few weeks.
https://www.youtube.com/watch?v=dFF4gN56Bhg
They actually don't have any video up of anything recent that actually really shows their core product. The video here shows a java client https://www.youtube.com/watch?v=5OYy_UtINo4 but it's from 2015.
Kind of a shame considering how much money and talent they've vacuumed out of the market. If I'm right, it's amazing for IBM to actually come in at a lower price than some competitor.