Obamacare "Tech Surge"

Great Debates
261

And that’s not going to happen.

These idiots still think that you can build, test, and deploy a software system in a month. If the payment and accounting systems were already code-complete they still wouldn’t ship by Jan 1 if any reputable engineering team was running the show. It takes at least that long just for regression testing, integration testing, load testing, installation, and acceptance testing.

What they will probably manage is some kind of hack job that insecurely takes payment information and forwards it to some mailbox and people will be hired to process them. It will be fraught with errors.

I can’t wait for the first claims to come in and it turns out that no one can find payment information for the policy. And any guesses how much fraud will be going on?

What kind of payment are they accepting, btw? Does anyone know? Many of the poor people this law was designed to help won’t have credit cards that are usable online, and they damned sure won’t have the kind of limits that allow them to charge a multi-thousand dollar health care policy.

This is just insanity - Payment processing is not a trivial task - especially if they have to set up monthly payments, handle chargebacks and bounced checks, and the like. They should have had that portion of the site being developed in parallel and that alone should probably have been a year long project or more.

262

I think they said the payments are payments to the insurance companies.

263

Yes, but what kind of payments are allowed? If they’re doing online payment processing, I assume that means credit cards and such. Or perhaps you can mail a cheque to an insurer, and part of the payment processing system is so that the insurer can notify the government that payment has been made so the customer doesn’t get fined.

That’s probably the next shoe to drop - come tax time, a lot of people are going to be fined who have paid for insurance but the system failed to notify the government. Or perhaps people will be marked as paid in the government systems while the insurers never receive the payment.

In the meantime, I’d like to see how they’re going to handle monthly payments, cards that expire, bounced cheques, and all the other myriad details of payment processing. Just being able to accept a credit card number, verify the payment, and have the money go to some account is just a small part of what a good payment processing system will do.

I assume other back-end software that still needs to be written will be things like management consoles, database administration, systems for printing invoices and mailing them, auditing software, diagnostics software, Reporting and query front ends and the like. None of this is trivial.

And they’d better do a hell of a lot better job on the security analysis and auditing of payment processing systems than they did with the rest of the project.

If someone told me that 30-40% of the system still needed to be built, and that included a lot of critical and complicated back-end functions, and said it would be ready Jan 1, my next question would be, “Do you mean Jan 1 2015? Or 2016? Because 2015 sounds a little ambitious.” I would not even consider that the person meant it would be ready in a month and a half. That doesn’t even pass the sniff test for being remotely feasible.

It sounds like the clowns are still running the show over there. They’ve learned nothing.

264

This gets more entertaining every time they reveal more of the details:

265

One interesting thing in there is that the govt forced CGI to use MarkLogic for the DB (it’s not an RDBMS, it’s a type of NoSQL DB).

1 - Not sure an RDBMS wouldn’t be appropriate for this project, if in charge of the project you would want to be pretty damn sure you needed to use something that is less mature than RDBMS

2 - If an RDBMS really couldn’t scale and MarkLogic or something like it was truly required, then it seems the prudent thing to do would be to build a prototype/framework to make sure the performance characteristics are understood before a full design is created

266

Bad managers ruined Obamacare (very interesting Nov/24 USA Today article)

" the politicians weren’t listening to the people doing the actual work"
“There is no ‘Suddenly Go Faster’ button, no way you can throw in money or additional developers as a late-stage accelerant; money is not directly tradable for either quality or speed, and adding more programmers to a late project makes it later.”

“Tech surge” is a political response to a problem … Not a reality based response.

“What engineers know that lawyers and politicians often don’t is that in the world of things, as opposed to people, there’s no escaping the sharp teeth of reality. But in law, and especially politics, inconvenient facts are merely inconvenient, something to be rationalized away.”

267

I bet they fix the website and things move forward and people gradually forget all about it. Well, everyone but you and yours.

268

Those of us in the industry will not forget this. Lessons learned are a powerful tool to avoid making similar mistakes.

269

Not really concerned with those of you in the industry.

270

Note to Obama: Never hire CGI again.

There, done.

271

Note to Obama: Now you understand what David Cutler was talking about 3 years ago.
This was Obama’s failure: not understanding how difficult these projects are and therefore not putting the correct people in charge of it.

272

Those of us in the industry already knew these lessons a long time ago. The problem is that the people running this particular boondoggle had no clue about the industry they were trying to manage.

The problems with this web site rollout were not just predictable, they were predicted. Most of us with experience in large projects like this saw this failure coming a long time ago. For years the Obama administration had been getting increasingly shrill warnings and pleadings from the tech side of this venture. They just chose not to listen. After all, who needs to pay attention to a lowly technician, when you’re surrounded by brilliant Harvard lawyers? Compared with running the world, how hard can a measly little tech project really be? So, the serious concerns of engineers were dismissed with the wave of a hand so the grownups could get back to what’s really important: politics.

And if you think the problem here falls on a contractor you know nothing, John Snow.

CGI, like QSSI and others, behaved exactly like contractors tend to behave when faced with intransigent idiots in management - they started playing ‘cover your ass’ games. They tried to raise the alarm repeatedly, and were met with indifference. When that kind of stuff happens developers start to become dispirited or indifferent, survival takes precedence over quality since you know you’re headed for failure anyway, morale suffers, and quality starts to go downhill. Blaming them for this failure is like blaming a horse for slowing down after you’ve beaten it repeatedly.

That’s not to say they’re blameless - it seems like there’s plenty of blame to go around. But you could have had the best programmers in the world on this project and it would have failed just as spectacularly so long as the same management team was in place.

Elvis, you’re an aeronautical engineer, correct? Have you studied the failures that led up to the Challenger disaster? There are eerie similarities. The engineers tried to raise warning flags, but there was such a large disconnect between them and the people running the political side of NASA that their warnings simply went unheeded - or not even understood.

The people making the ultimate decisions were so incompetent on matters of engineering that they couldn’t even understand the language that the engineers were speaking, and so used to living in a world where problems could simply be negotiated away that they didn’t respect problems of fundamental engineering which won’t go away just because you really, really want them to or because you’re really clever at hiding them or deflecting them by giving a great speech.

273

One of the things that I keep scratching my head over is that they didn’t start coding this thing until spring of 2013, and using a technology they were unfamiliar with (MarkLogic). Spring of 2013 for an Oct 1 go live for a very complex project. Just absolutely clueless.

It’s so profoundly clueless that I really have lost a vast amount of respect for Obama (I voted for him). He may not have experience with projects like this but part of the due diligence when planning/launching the project is understand what you are getting into and putting the right people/processes in place to ensure success.

274

Yeah, that seems to be the attitude that got us here.

275

What I suspect happened is what often happens with first-time managers of large, complex projects - they begin by choosing the deadline, and plan backwards from that.

Regards,
Shodan

276

It’s not like they had a choice. The administration was legally bound to do certain things, and there was zero chance any of that would change.

All in all, for a project who’s scope dramatically increased (remember, states were supposed to be the ones setting up the exchanges), having the site pretty much working two months late is not bad.

277

Oscar Wilde once said that whenever a person did a really stupid thing, he usually did it for the very best of reasons.

Plus it’s a little funny-sounding to hear that the Obama administration had no choice about a law that they conceived, pushed thru, and were responsible for implementing.

Regards,
Shodan

278

And regularly make changes to it whenever they feel like it now.

279

Are we blaming the Obama administration for not being able to predict the future?

The scope of the project dramatically increased when states refused to set up exchanges for political reasons. The administration couldn’t realistically change deadlines or allocate more resources to account for the increased scope due to opposition in Congress. It’s obvious what happened. Republicans wanted ACA to fail and they did everything they could do to torpedo it. As I said before, the fact that Healthcare.gov is pretty much working 2 months after it was supposed to is a pretty good showing.

280

Wether that was for “political reasons” or not, it wasn’t rocket science to expect many states to opt out. That’s not predicting the future, that’s understanding the implications of your policy.

Please. They’ve done that several times recently.

Obama didn’t need permission from Congress to fix the problem, and he didn’t need permission to do it right.

It’s obvious what happened. The Obama administration fumbled the ball, as Obama himself said. Republican opposition to the law was nothing new, and it was the Democrats choice to pass the law with 0 support from Republicans in the first place.