back
93 comments
A bunch of the folks in this article are Y Combinator or Google alumni, and we're giving a talk on Saving HealthCare.gov next Thursday in SF for anyone who's interested:

http://www.eventbrite.com/e/saving-healthcaregov-tickets-118...

Let us know if you have questions! Or if you want to help.

UPDATE: we seem to be sold out. Trying to find a larger venue. In the meantime, feel free to sign up on the waiting list and you'll get an email as soon as we find a room with more capacity.

Not to start a political flame war, but...

How do you, as technologically savvy people who are in a better position than most to understand the grave transgressions of the NSA, reconcile your support of an administration that has shown complete disregard for our privacy and civil liberties (specifically within the scope of the NSA/Snowden scandal) with your inherent duty, as an engineer of powerful systems, to above all further the public good? As a subtext, what are your thoughts on continuing to further said administration's political aims in one arena while they restrict your rights and livelihood* in another?

Furthermore, and more on topic, what steps have you taken to place privacy concerns at the forefront when creating a service that consumes vast amounts of personal and medical information from millions of people?

Last, as someone who has been "on the inside" do you see corruption as a large problem with government procurement? To what extent did corruption create the situation that you were called in to fix with the original HealthCare.gov fiasco?

*assuming, of course, that the NSA/Snowden revelations have hurt US tech interests

Edit: Mods, if you disapprove of this comment feel free to delete it. That said, I think these are some important questions that need to be asked.

You can both support building a tool to give people healthcare and be against government surveillance. Most people are able to compartmentalize their support or opposition for various government policies and do not see the doing of one thing they don't like as a reason not to support any other thing the government does.
If everyone tries to reconcile everything the government does, no good person would ever work for the government.

If you want to change it, change it from the inside.

Thats is what they are doing.

People can protest on the streets or online but changing the government from the inside will be more effective.

While I can see huge ethical concerns were this project related to the NSA, I don't think you can necessarily force that association here. While I agree with your frustration with the administration, especially with regards to surveillance and civil rights, I also would like to see the ACA do well and lead to a better life for most Americans. I don't think its wrong to want to work on solving this problem (healthcare) while disagreeing with other actions of the government, as the above poster is working on something that will directly help millions of americans. I see improving healthcare as a separate issue from the NSA scandals, and I think that, in refusing to help with fixing healthcare.gov, you are harming millions of poor and needy Americans without actually accomplishing much of anything.
I'm not the person you asked, but there seem to be quite a few things wrong with the assumptions underlying this question.

> an administration that has shown complete disregard for our privacy and civil liberties

Is it really true that this is the case for this administration more so than other administrations? Admittedly the increase in opportunity for surveillance has been huge in the last couple decades so the quality of the comparison degrades as you go further back in time, but what the NSA has been doing seems entirely in line with what they've been doing for decades. To be clear, what they've been doing for decades is really, really shitty and I'm glad people are finally outraged, but then wouldn't your view require basically boycotting working with any administration ever? I don't think it's an unreasonable proposition to not hold hostage policies that are potentially worthy just for a single (however significant) issue that you disagree with.

> your support of an administration

> continuing to further said administration's political aims

This leads to me to my next point: it seems like your comment assumes that this was done purely out of a desire to help a political figure, instead of considering the possibility that the volunteers may think that Healthcare.gov is a worthy goal or at least that its brokenness is negatively affecting the American people (I don't think there's any disagreement across the political spectrum on this latter point). That's an awfully cynically political way to look at the world, and I don't think it's accurate to assume that that viewpoint is shared by the volunteers. I think it's probably a safe assumption that the volunteers acted more out of a desire to see hc.gov working than to see Obama's political capital enhanced per se.

Do you quit your job when your boss does something you don't like? The government will live on beyond this administration and hopefully the site will as well. I don't know what privacy you can expect when they hold the keys, my expectation would be 0. Finally, government procurement is always littered with inefficiency and occasionally large scale corruption. These questions really answer themselves.

Now if only you had asked something about technology.

Here is something that'll blow your mind: lots of technologically savvy people don't think surveillance is a policy so bad that it destroys the governments overall credibility. Some even don't see a problem with it.
If you're starting from the assumption that everyone else who is "technologically savvy" is obviously in agreement with you, then you're making a large mistake in your reasoning somewhere.

Smart people often disagree about things in this world, including in particular whether the actions of the NSA show "complete disregard for our privacy and civil liberties" and whether they're justified or not.

Also, half the idea of democracy is to compromise with other people. Thinking that people shouldn't do any work for the government because they don't agree with some of the actions of that government is ridiculous.

To make the coupling explicit, what assurance do we have that information given to HealthCare.gov won't be illegally accessed by employees at NSA and other three-letter agencies?
But how would you ever make government better without getting involved? I mean unless you are talking about scrapping the existing system and replacing it, you must be involved in order to change it.
Pretty much everything you wrote is complete nonsense. There's no relation between healthcare and spying. People don't put their medical information into the system (hello! no more denials for pre-existing conditions). "Corrupt"? Really? Then you try to shame the folks for participating but even acknowledge that inanity with your caveat. Pretty shameful all the way around.
Did you redesign it from scratch or were you able to leverage any existing code?

Were the original developers developing against a (mostly) static spec, or were the terms constantly being changed up until October of last year?

Are there any plans to carry this forward and redesign some of the state exchanges?

Mostly from scratch.

Cannot speak directly to experience of original developers, but I could see moving target as being an issue.

Would love to see our code open sourced someday and be useful to others, especially state exchanges, but that is a long-term goal. Currently we are laser-focused on getting everything ready to ship by next open enrollment.

Even if you don't release it until after a successful open enrollment, please record video! And please, during Q&A, remind the speakers to repeat the question so it's captured on audio! Or capture the audio from the active audience microphone!
Why does it matter if some of the people are Google alumni? That doesn't mean anything anymore, the company employs over 10,000 engineers. It's like saying that you have IBM alumni.
I see it's sold out now. Will there be any possibility of a second event or will the rest of us also be able to find out how we can help after the event happens?
Yes — we're trying to find a larger venue. For now, sign up for the waitlist and you'll get an email once we've found something larger!
Could you please consider posting videos of the talk? A local group on software engineering did a talk I could not attend but I am interested.
Any chance we could live stream this in Boulder?
Unfortunately we're not doing any recordings -- it's hard to be candid and recorded for all history at the same time. :(
Damn, what luck. Gonna be in town just for next week, I definitely must attend this.
Are these devs working on healthcare.gov being paid or volunteering?
Are Tor users going to be blocked from accessing the new site also?
So cool! Just got my ticket.
There are plenty more government IT projects for them to save next. The scale of these failures is mind-boggling, and Healthcare.gov wasn't nearly the worst. Here's one I just read about yesterday, might be good for their next project:

http://www.thefiscaltimes.com/Articles/2013/11/26/Another-Fa...

That's $1.1 billion lost, another $4 billion planned. I can't even imagine what I'd demand of a $4 billion piece of software.

If you're interested in helping with this or any other government project, feel free to email jobs@hcgov.us or attend the talk mentioned elsewhere in the thread! There are lots of ways to get involved.
I can't find a reference for this, but I've heard that the IRS has attempted several multi-billion-dollar system overhauls, with no success.

It's fascinating how much of the world is run by decades-old software, with no serious hope of an upgrade.

The question I have is, how do we get government to hire teams like this in the first place?
Build a federal gov run IT branch that isn't the NSA instead of contracting things like this out?
It's incredibly difficult. When government money is involved the good old boy network swings into action. Contractors are picked based on political connections first and actual capabilities second.
So... are we going to get any of our taxpayer money back from the people who fucked up the first time?

On another note I am really delighted to see that the government was forced to step out of their bubble to make this happen.

Nope, in fact we're still giving them money for new projects. http://www.cgi.com/en/media-announcement
This is a very important issue that failed in our last attempt.

As we put so many resources behind that attempt, this proves it's a difficult problem to solve that requires more resources.

It logically follows that more taxpayer money is required for the next iteration. How else could we solve such an important problem?

Could you imagine if every shitty software project ended in a refund?
We built a NEW website, and it's going to WORK this time. Not like that last time, that only took us up 30 feet into the air then dropped us. And when it DOES work, you're gonna turn to me and say, "Dude, this rocks." And I'll be like, "Dude, I TOLD you it was gonna be sweet."
Why is this using Amazon servers? Wouldn't it make more sense to use purpose-built machines for this?
No. You serve the static assets from a CDN, but the authentication and dynamic content generation need to scale very quickly when you have an in rush of traffic. This is the perfect use case for Amazon AWS (GovCloud, specifically).

EDIT: Disclaimer: I did not work on Healthcare.gov, but did study up on the entire architecture while trying for a gig on the recovery team.

The government has their own amazon cloud, but to the point I think it actually would be better to use a virtualized setup because then you can scale power as needed and not have to worry too much about bringing servers online. Then when the inevitable slow period comes you can scale down significantly. This is my experience at least.
I hope they're performing load tests this time, which apparently was not done properly for the first version.
Talk is cheap.
> David Chang, a 2102 MIT graduate

So that's how they're doing it, they got a programmer from the future to help!

Well the US federal government has more resources than most so it strikes me as a smart move to do this. I suspect they were holding off announcing that they've solved time travel closer to the congressional elections to help out the Democrats but the cats out of the bag now.
How do they measure success? If something costs 1000x what it would have in the market and still strives for adequacy it's already proof of failure. If any of us failed so spectacularly we'd go bankrupt and another service would be used (and there would be plenty of choices if regulations/poor law services didn't dissuade people from creating this kind of service).

Now we have people that were productive at companies making services we pay for of our own volition spending time on this thing that doesn't work as a symptom, not a cause. So confused why this is a thing.