Audit: <a href="http://budget.nv.gov/uploadedFiles/budgetnvgov/content/IAudits/About/AuditRpts/C18-01%20DMV%20Audit.pdf" rel="nofollow">http://budget.nv.gov/uploadedFiles/budgetnvgov/content/IAudi...</a><p>This one is extra bad. Not even a communication plan? They did manage to spend $13 MILLION on Oracle shit though. The best part is the solution they sold, which is not even working in New Hampshire, and was built on DynamicsCRM (not portable to oracle)<p>The grand finale:<p>"The contract was amended in February 2017, which extended the termination date for an additional year for maintenance coverage. It also increased the maximum amount of the contract by $3 million, primarily for additional hardware and software."<p>That person needs to be fired immediately and probably arrested.
It's kinda crazy how much money gets thrown around by an org you wouldn't think twice about. If a group of engineers from HN got together and secured a deal like that, they could easily retire with that kind of money.
Prior version of the system used PowerBuilder. That figures.<p>My fly-by of PowerBuilder 20 years ago made a deep impression. Like, I never wanted to even <i>think</i> about that hunk of garbage again. It's right up there with PeopleSoft. If you're ever offered an opportunity to work with these systems, just run away.<p>The state didn't know how to buy or develop software then, and it doesn't know how to buy or develop software today.
I wish code for america would take these kinds of things on. I assume the DMV rules are sufficiently similar that most/all of them could benefit from opensource.<p><a href="https://www.codeforamerica.org/" rel="nofollow">https://www.codeforamerica.org/</a>
How does this happen?<p>Who in their right mind thought it was a good idea to outsource the development of this critical system to a foreign contractor? Did anybody stop and think about what might happen to all of that public money if the contract is breached?
According to the audit document, the company was Tech Mahindra.<p>I am genuinely surprised when a 8 to 9 figure software project ever succeeds.<p>But I'm also amazed that the government spent $25mil before they realized they got completely bamboozled. There's something shady going on here.
$78m over 5 years for 25 (probably average) developers and a few licenses is excessive. Not sure what massive tech problems the Nevada DMV is trying to solve, but I doubt it's much more than a DB cluster with a web app or two thrown in front of it. Tech Mahindra's developers may be average, but their marketing department seems to be exceptional!
This type of waste occurs in all levels of government. My last employer was spending over $1,000,000/year maintaining some horrible SharePoint sites that were rarely used.
I'm sure that money will find its way back to America, in political donations to be exact. Kind of gullible to think these 'failed to deliver' accidents keep happening on accident after half the budget has been spent.
I would love to see how much of the purchasing and decision making committee was composed of people who didn't understand technology, let alone how to procure and implement technology.<p>At the highest levels, there is a self-preservation disease permeating most organizations where CxO's try to fake it till they make it with the IT/Systems decisions they take on, and it turns out like this.
Ah the good old responsibility dodging and blaming following legacy modernization project gone off the rails. At least they got out spending only $27 million. The state says bait and switch, I wager the contractor never got requirements or basic documentation for the project.<p>I’m at a company now that’s planning one of these projects for a core system. I’m looking forward to having a front row seat!
Tech Mahindra contractors, not surprising. An acquaintance of mine came in on a job with them- they pay all their guys salary which is kind of crazy. That guy was doing better than the hourly contractors with family time and pay. Something sketchy is up with them.
I am currently working on a largish COBOL codebase which the company has tried 4 times to port to Java, each time by porting all at once. It cannot be done all at once.