Thursday, December 12, 2013

The Big Data Job Interview

I had the pleasure of interviewing with one of the top firms in the Big Data space yesterday.   Since the firm produces a wide variety of products, our discussion quickly turned to which section of the company I was interested in.   "I'm passionate about big data," was my initial response, "but I'm also quite experienced in mobile and database technology."

A little bit later in the interview I was asked a question that I had anticipated.   "I'm a C-Level executive.  How would you sell me on Big Data?"

I stalled for a bit of time, and tried to highlight my flexibility and confidence with the subject matter by countering with a question of my own.  "Which industry?"

The interviewer was an actual hiring manager, a VP of Sales too experienced to let me wrangle a bit of control of the process by allowing me to be the one asking the questions at this early stage.   Without hesitation I was countered with the simple but very effective phrase, "You pick."

I went with the Healthcare industry because of HIPPA requirements in the US and because the obvious process optimization gains hospitals would be able to realize by identifying both positive and negative outliers, adopting the positive and eliminating the negative, on their way to establishing new best practices going forward.   I outlined briefly that patients could be considered "Work In Process" and that operations could be sped up and made more secure by establishing patient identity cards, containing biometric security verification, retinal scan or fingerprint id, thereby ensuring that the patient is in fact who they say they are.   This would eliminate a lot of fraud from the system early on.  Coupling this card to an RFID tagged bracelet, would allow the hospital to track the 'product' through the entire process, collecting more data along the way.  

"For what purpose?"  was a confirmation question I faced at that point.   

Patient safety, making sure that the patient who was supposed to undergo a surgical procedure, was the one undergoing the procedure.   Also ensuring that no transcription errors occur and that the patient receives exactly the correct drug and dosage amounts prescribed, as opposed to problems that can arise from other medical professionals or pharmacists trying to interpret the doctor's handwriting. So you could say by beefing up their systems, C-level executives would be able to mitigate risk moving forward.  With the gradual integration of the "Internet of Things" or intelligent medical devices and sensors, more and more data could be collected as more and more steps of the process are electronically or digitally monitored and automated.

Beyond data collection, you can now apply analytics.   This can allow you to determine optimal procedures for certain ailments.  Is surgery required or is drug therapy the optimal solution?    Do the surgeons working for your hospital have areas of specialty?   Can you allocate cases more optimally along these lines?   Almost like a baseball manager, the proper system would be able to predetermine the hospitals optimal 'batting' lineup based on their workload that day.   The net result would be more patients served, more safely, and by the best person capable of doing the work 'at the plate.'   For the hospital it would mean serving the patients with more optimal care,  at statistically, the lowest possible costs, while enhancing rather than compromising quality.

I then touched on mobile.   The advantages of having medical information available on tablets for doctors and nurses alike.   The security of medical information on devices is far superior to a clipboard at bedside, but it also allows specialists to consult on cases remotely, with full patient information, perhaps even live vitals feed, whether they are offsite at dinner or on the golf course.  Tablets themselves are perfectly suited for the hospital environment, because they are the same form factor as the clipboards they would be replacing, which would permit a direct digitalization of forms currently being used and a much shorter learning curve.   Tablets are also much easier to keep sanitary then computer keyboards or a paper-based system. 

Last, I touched on the monetization of data over time.   I cited examples of where some drug companies were willing to pay for patient outcome data to validate the effectiveness of their products.   Statistics on procedure effectiveness, drug effectiveness, wait times, population demographics, could all be very valuable to third parties.  But first, the hospital would have to be able to reliably collect and then 'package' the data properly.  HMOs and governments would be able to utilize the same data and with predictive analytics, allowing them to refine their capacity planning going forward.   So in short, big data and accompanying technologies, would be able to increase revenue, reduce costs and maximize the utilization of their current assets.  What's not to like?

I thought I had done well on the interview.   But then on the drive home, self doubt started to creep in.   Had I got too bogged down in technical details?    I resented not hitting that home run catch phrase, that would, differentiate me from the rest of the field.

It was dark on the drive home.   I was stuck in highway stop and go traffic for more than an hour.   My mind automatically kept going over what had transpired in the interview.   My dream job on the line and I had put in an okay, but not spectacular performance.   I began to think that what I lacked was a great universal analogy to describe the value of big data to any level of business executive.   An analogy so effective that once used in an interview, would be recognized immediately as a home run, gaining me a job offer on the spot.   As traffic started to finally approach the posted speed limit,  I spotted an opening, and at least was able to separate myself from this pack. 

A few exchanges later, as I left the metropolitan area and my grandiose dreams of the perfect big data analogy behind, I found myself  now alone on what appeared to be a very long stretch of highway, at night.

So I did what any other driver would do in these circumstances.  I turned on my high beams.

And then it hit me.

A C-level executive is really like the driver of a car, coping with ever increasing speeds and huge areas of uncertainty, or darkness.   Dashboards are of little use to that level of executive.  Dashboards are merely a reflection or validation of what's happening right now.  Today's results are merely the active history of decisions already made long ago.  Looking down at the dashboard isn't going to help the C-Level executive determine what lies ahead.   What the C-Level executive needs is something that will help enhance their vision of the future, or extend their vision of the road ahead at night.    

Having access to big data, would be like turning on your high beams, giving C-Level executives that valuable extra time required to make very important decisions ahead of time.   To take it one step further, predictive analytics could be considered an extra set of fog lights.  Having access to the two in tandem would allow any C-Level executive to avoid costly, if not life-threatening accidents, in the process.  Given similar conditions of darkness and speed, a car with high beams, or a company with big data, would be able to out maneuver any competitor no so equipped.  Both high beams for cars, and big data for companies, allows one to better navigate through all sorts of nasty weather or business conditions.  Both extend one's vision ahead into the future when it's needed most.

As I turned off at my exit, and reduced my speed to adjust to residential speed limits, I also turned off my high beams.   For the rest of the way home I tried to come up with a way of creatively getting this big data analogy through to the hiring manager, without appearing too obvious. 

Does anyone out there have any big ideas?

"Vision is the art of seeing what is invisible to others."      - Johnathan Swift

No comments: