Passing around financial data on SFTP drops is something that irked me. Having the total not add up from the line items is also pretty ridiculous. The systems in place are all about edge cases that have formed over decades.
Software engineers are not invited to the discussions in a lot of these companies. The area is gate kept by healthcare people, which tend to be people who don't know how to do technical writing or planning.
Our company wasn't a huge one. But the huge ones looked to me as run by accountants. If you can't establish an income stream for them, they won't even talk to you.
I see this a moderate amount in fintech, or perhaps at least older fintech?
Had a colleague tell me about a new financial info provider he was supposed to connect with for data. "Oh this one will be easy" he was told. "They use oauth". OK... well... you oauth in, then you hit an endpoint that triggers a data build which then puts your data file in to an FTP account of your configuration. But you can't even configure it via hitting the endpoint behind the oauth. You have to log in to a web app to configure the FTP account for your account.
You hit the API endpoint and get 'OK'. But... that's it. You might get a file in your FTP account 5 minutes later. Or an hour later. It's just whenever it gets processed on their end. And... if it doesn't show up - you have no indication whether something broke, or perhaps the data service "just doesn't have any data for you today - you can't expect data every single day".
It's insanely crazy how across-the-board poor so many of these mid-level data vendors are. Colleague is currently supporting 8 different integration ingests like this - some deliver FTP daily, some you have to pull down, etc.
One delivers data which is only ever a delta. You have to continually request previous daily info from them to get to the 'first record' for that account, which can sometimes go back years, but... you don't know that up front. So you have to loop and request "day before", pull down data, parse it, then check the day before, until you don't get anything for that customer, then presume the previous was "initial balance". But you also can't just do that once, because... there may not have been any data the day before because... holiday? system down? So in practice you need to go back 3-4 days at least to verify. Oh... and don't keep looping and requesting too fast... you'll exceed the low rate limit ("we don't want people abusing our API").
I can only barely imagine how systems evolve like this, because they don't really seem to serve anyone's interest fully.
And btw, that description given by your colleague brought cringy mental flashbacks on meetings I've recently attended.
I also worked with a state medicaid office. They refused to use DNS, they decided it was insecure. Which was lots of fun any time we changed datacenters (3 times over my 7 years there). They would also go down, and nothing would happen until around 9:30-10am the following work day when they would finally come around and fix it.
The most competitive/efficient markets generally need the optimal amount of regulation to keep/get them that way. Those regulations would ensure that these conditions continue to be met in order for it to be a competitive market: https://en.wikipedia.org/wiki/Perfect_competition#Idealizing...
Free markets that are completely unregulated will tend to veer towards collusion and monopolies, in which case they're no longer competitive and the customers suffer. Free markets that are overly regulated will also suffer, but more-so due to things like the regulatory burden/barriers for new entrants.
What we have in the US healthcare system is a demented hybrid where some parts are free market but under-regulated, while other parts are free market but over-regulated. It's the worst of free market and state-run combined.
But in reality, good regulation is good, and bad regulation is bad.
It sounds silly written out because it's a tautology, but if more people internalized it we could have much more productive conversations. Instead of going back and forth on whether regulation is good or bad, we should try to figure out which regulations are good and which are bad, and then try to get rid of the bad ones and pass more of the good ones.
No one brick in the road to hell is bad. But you can definitely pave a road to hell with the accumulation of "positive" changes. People tend to define positive over too short a timeline or too narrowly.
They're just rules we make for ourselves. There's nothing special about them. Opinions like that are why we can't fix messes like this.
Perhaps its just low pay attracting low talent
The goal of the regulation is to increase prices (profit) while preventing other parties from competing on prices
Suppose I want to sell my product, and find a company willing to buy, a middle manager signs the contract. When the company recieved my product, it refuses to pay, says this person was not authorised to sign the contract on behalf of the company.
Must the company prove the purchase was unauthorised, or must I prove the employee was authorised? Suppose employee was unauthorised, is it my or the company's problem if this guy 'took initiative', who eats the loss?
What if they already used my product and only realised after the bill came due? What if the deal was overpriced like 10x from market average? What if the guy's title says 'Head of procurement' but the company swears their bylaws don't allow him to authorise purchases?
And I think you're talking about contract law, not regulations.
(Source: was doing medtech development for a while in the early 2010s).
I don’t know how it works in the US but at least in France, most hospitals are part of universities so they are also doing a lot of research.
It's precisely the opposite. Its a racket but its not for profit. If you made a health care service that didn't have any of this cruft and just provided great care, it would not be used, because it would not be compliant with insurance, and employees would not get the tax benefit of using it.
If it were for profit, it would work better.
https://www.cigna.com/employers-brokers/insights/informed-on...
(Germany has private insurance and France as well in parallel, but the main system is "sozialised")
The healthcare systems in Canada, the UK, France, Germany and Switzerland are all distinctive (they’re main commonality I suppose is that they somehow deliver quality healthcare at a fraction of the US cost.). But to hear Americans say we should “do it the way it’s done in Europe” is to understand that the person likely don’t know how the healthcare systems operate in different countries.
The other frustration in talking about this is the tendency to “It’s just” the problems. If this were a simple problem with an obvious solution it likely would have been solved. We are where we are because of a century of organic, and often idiosyncratic, growth. Well intentioned policies of decades ago have come back to bite us with unintended consequences as likely are our best ideas today.
I don’t say we should throw up our hands and give up, but when anyone says there is an easy solution - we just need the political will to let them do what they want - I do see snake oil.
"Some of you may not be in compliance. We start you off at a fine of $1. For every month this continues, the fine doubles."
All you need for a sensible regulatory environment is a complete overhaul of how your society works, starting with a campaign finance & congressional staffing system that isn't built on implicit corruption.
Fine these people out of existence and let someone competent (if such a person exists) pick up the pieces and start anew.
Eh, I'd rather listen to duffpkg, with his extensive, real world experience in health tech, than a random comment.
FWIW, a former coworker of mine spent a few years at a health technology company (also worked with MUMPS), and legal compliance was the number one hurdle to modernizing. It was expensive, and created a huge drag.