The Big Data Talent Shortage: Are H1-B Visa Holders the Solution?

There are recent arguments claiming that there is not a tech talent shortage in the US based on a surplus of US graduates with Science, Technology, Engineering and Math degrees and stalled salaries for the past decade in these same areas. The study by the Economic Policy Institute points to the laws of supply and demand: if there were a shortage, the wages would be rising in these fields. But critics argue that immigrants holding H1-B visas are partially to blame for salaries being at a standstill. It did recognize that there might be a shortage in the harder to fill specialized areas. (Cloud Developer, Data Scientist, App developer, etc.)
By design, the H1-B visa program allows immigrants with a Bachelor’s degree and a highly sought after skillset to be hired by companies and be ‘sponsored’ for 6 years only if there are no US Citizens or green card holders qualified for the job.  If they graduate and do not find a company willing to sponsor them, they are forced to either leave the country or go back to school. Sponsored workers also have to jump through legal hoops to transfer to a different company within the 6 year period so their retention rate is higher compared to the retention rates of non-sponsored workers.
If a company says ‘we’d like to make you an offer and we will sponsor your visa’, do you really think the recent graduate will negotiate the salary much? Is it also possible that companies would rather hire a cheaper, foreign worker than an experienced US citizen commanding a higher salary? But think about another variable. What if the new graduate was more qualified for the analytics role straight out of school compared to someone with 15 years of experience?
Certain Fortune 100 companies claim that the US is hobbled economically due to the lack of tech talent and lobbying to increase the number of visas being issued to immigrant workers. The number of H1-Bs available to for-profit companies is currently capped at 65,000 but the new Senate plan proposes an increase to 110,000 to 180,000.
It has been highly chronicled that the Big Data talent shortage is turning into a storm of epic proportions and the number of Big Data Analytics positions is expected to reach 4 million globally by 2015. The newly released Big Data Jobs Index powered by icrunchdata shows the top 6 specializations in the Big Data space and the number of positions currently open. Topping this list are 217,346 Analytics positions currently available.
Universities are quickly adapting and redesigning their math and statistics programs to better align with the requirements in data science and analytics. Williams College, Harvard University and UC Berkeley are all reporting enrollment increases in statistics-focused degrees.
So will H1-B visa holders be the solution to the Big Data talent shortage? Yes, if a few things happen. More university statistics departments will have to evolve their programs to match the unique requirements in Big Data and the number of available H1-B visas to companies will have to increase. If experienced tech workers do not seek training and schooling in the ever changing analytics space, they will be replaced by H1-B visa holders coming out of school already more qualified. The immigration report states that 36% of IT workers do no hold a college degree and only 24% of IT workers have a four-year computer science or math degree. But the stagnation of salaries might not be repeated. If older workers fail to get trained on newer techniques, then fewer, albeit younger people will be filling those rolls. But, if fewer people qualify for these positions, the law of supply and demand will kick in and the salary levels will increase.

Original Source : http://smartdatacollective.com/todd-nevins/122981/big-data-talent-shortage-are-h1-b-visa-holders-solution

Read More

MicroStrategy 9.3.1 Free Evaluation Software Download

 

For All BI Applications from Workgroup to Enterprise

Download Free Evaluation Software Today! MicroStrategy helps thousands of customers make better business decisions every day with unified Dashboards, Reporting, Analysis, and real-time business Monitoring. To see it for yourself, give MicroStrategy's free Evaluation Edition a try. Now available for both Windows and Linux.

Download Trial: Windows

Note:Postal shipments of MicroStrategy Evaluation Edition Version 9 Release 9 Software will not be released until 05/10/13 at the earliest. Downloads of MicroStrategy Evaluation Edition Version 9 Release 9 Software are available immediately.

Read More

PriceHub Wants To Tell You How Much Your Car Is Really Worth, With Data To Prove It

How much is your car worth?
It’s an easy enough question to answer. Punch in the details at Kelley Blue Book, hit submit, and bam — question answered, ego stoked (or not.)
But how do they know how much it’s worth? For the most part, even the tried-and-true sources like the ol’ Blue Book are kind of a black box.
PriceHub wants to make the process more transparent. They’ll tell you how much your car is worth, and give you a mountain of data to back it up.
Update: Whoops! Looks like we crashed their server. The PriceHub team says they’re working on getting it back up ASAP.
Like many a car valuing service, PriceHub pulls its transaction data from all sorts of sources. Some of it comes from users; some of it comes from dealerships, or used car auctions. The vast majority of it, says the company, comes from DMV records.
Unlike most other services, though, PriceHub makes a ton of this data available directly to the user for their own perusing. Want to see the transaction details for 10,000+ Honda Civics sold in the past 18 months? Sure. Want to limit it to just 2009-model Civics sold in California? Hey, why not.
PriceHub actually came into existence with little to no fanfare a few years back, built as something of a hobby project by Myron Lo, then the VP of Innovation at ZipReality. It lived the first few years of its life in a rather humble form; black text spilled across the white background, with a modest data set of around 50,000 transactions.
Over time, however, it became clear that this lil’ pet project could be something more. As the site naturally grew toward its 50,000th registered user, the team behind it decided to dive in headfirst. They applied to Adeo Ressi’s Founder’s Institute, got in, and have spent the last few months being “whipped into a start-up by Adeo” (their words!)

In that time, the team has added all sorts of new tricks to PriceHub’s repertoire:

• The old, Geocities-tastic design has been overhauled into something significantly more modern
• They’ve bumped their data set up 10X, from 50,000 records to 500,000
• They’ve added model research records, bringing in things like recall alerts, safety ratings, cost estimates, and service bulletins.
• They’ve added depreciation charts for each model year of a car, giving would-be buyers/sellers a rough idea of how quickly the car in question is losing its value
• The team has grown from one to three: the aforementioned Myron Lo, as well as Telly Chang (former Product Marketing lead at Yahoo! Autos) and Sandy Lo (currently also Marketing Lead at the Salesforce-backed Financialforce)

Speaking of depreciation, the company mentioned a work-in-progress feature that I find particular interesting: depreciation alerts. If you’ve told PriceHub that you own a certain car and their data starts to suggest that its value is startin’ to turn, they’ll soon be able to fire off an alert to let you know that it might be good to sell sooner than later. Also on the roadmap: mobile apps (of course), and Zillow-esque sale price vs. time-on-the-market data.
Anecdote time! Around half a year ago, I sold the first car I’d ever owned: a 2002 Honda Accord, which I’d more or less driven into the ground. I sold it to the first person with a stack of cash and a pretty smile, letting it go for a bit over $2.5 grand. According to PriceHub’s records for cars of that year with similar mileage, I probably could’ve gotten another two thousand bucks out of the car if I’d been patient. Whoops!
PriceHub isn’t alone in this space, of course. Transparent or not, legacy offerings like Kelley Blue Book and Edmunds have held the throne for decades, with relatively new folk like TrueCar chipping away at their lead for a few years already. What do you think: is data and transparency enough to make PriceHub standout?

Original Source : http://techcrunch.com/2013/05/06/pricehub-wants-to-tell-you-how-much-your-car-is-really-worth-with-data-to-prove-it/

Read More

Warren Buffett, the Human Big Data Engine

Warren Buffett is one of the world’s most successful investors. By combining a sound investment philosophy with a bullet-proof decision-making process, Buffet has been able to outperform the stock market by roughly 13% over a 35-year period.

Buffett spends countless hours researching each of his equities. Before making a decision, he reads annual reports, news publications and any other information he can get his hands on. His decisions are informed by research, and based on the strict fundamentals that work best for his style of investing.

The way Buffett operates is not unlike the best use cases for big data. Among other things, his success lies in his ability to make good decisions in accordance with quality benchmarks. This is the key to using big data well, and to business success in general.

Big data users can learn from Warren Buffett. His brain, in effect, is a kind of big data engine. Here are three lessons that Buffett provides for big data users:

1. Know What You’re Looking For

By analyzing large amounts of data from diverse sources, we can gain context and correlations that wouldn’t have otherwise occurred to us. Buffett pulls from actuary reports and bond publications; big data can suck information out of nearly everything, from social media to expense reports.

But it is all for nothing if you don’t know what you’re looking for. You need structure around your queries.

Buffett has a few rules of thumb. He looks for low-volatility stocks with a low price-to-book ratio. They should be profitable and growing, among other considerations.

Similarly, if you ask your data analytics platform the right questions, you’ll know the best answers when you see them. Perhaps you’re looking for a Scotch-drinking customer base that prefers premium product and prioritizes convenience over shopping around. If you know that before you ask your data engine to generate their social media preferences, you’ll get a much better result than if you asked the same question about Scotch drinkers in general.

2. Understand How the Object of Your Research Works

Buffett once said: “Never invest in a business you can’t understand.” Given Buffett’s rational investment style, one can surmise the meaning behind that quote. If you don’t understand the mechanics of your investment, you’re relying too much on emotion and guesswork. Or else the business itself isn’t viable.

Likewise, if you don’t understand how the organization or customer base you’re running a big data query on works, you’re relying too heavily on assumptions. You could ask questions of the data that have nothing to do with the reality of the situation. Asking questions about the rate of email responses in a customer support organization that relies mostly on forums won’t get you very far.

3. Use Solid Benchmarks

Some of Warren Buffett’s more famous quotes illustrate the precision and simplicity of his benchmarks:

• “Rule No. 1: never lose money. Rule No. 2: don’t forget rule No. 1.”
• “Only buy something that you’d be perfectly happy to hold if the market shut down for 10 years.”
• “It’s far better to buy a wonderful company at a fair price than a fair company at a wonderful price.”

These benchmarks demand the best. They don’t compromise. Arrange your own benchmarks to similar standards, ask the right questions, and make data driven decisions that reflect your high standards and style.

Make Data Your Servant

Maybe one of Warren Buffett’s biggest secrets to success is that he made the data work for him. If you aspire to do the same, data will quickly become your best friend.

Author: Chanu Damarla, Senior Director of Product Management at GoodData. 

Original Article :http://smartdatacollective.com/andreesa/122086/warren-buffett-human-big-data-engine

Read More