Back where I started
Occurs to me that my present is in some ways not different from my past. :)
Recently posted a take on putting together a cloud. There is a lot of educated guesswork in such an exercise, as there are a lot of unknowns.
Back when in college, I could not afford to go directly to University, so spend a couple years working. Started working at a ratty local bike shop, and was later recruited to work at perhaps one of the best in the country. (I was ... a bit intense as club rider. Later found my training buddies were UCSF category 2 racers.) The shop ("Two Wheel Transit Authority" in Huntington Beach) specialized in high-end touring bicycles and gear. Imagine as a 19-year-old kid, advising customers on what in current prices is $10K-30K worth of custom-built bicycles and gear. Further imagine those customers are going to spend weeks or months in Europe, living with the gear you recommended.
Thought at the time: If I get this wrong, they are going to hate me.
Know further that none of this was exact. What is their current level of fitness? What is their character (easy-going or hard-chargers)? How much experience did they have with bicycle touring? What sort of terrain is on their route (mustly flat, or serious mountains)? On a strictly deterministic basis, I was making wild guesses on entirely insufficient data.
Yet somehow, it worked, and worked well.
One of my customers walked in with a department-store bike, and asked my opinion. Told the exact truth - it was unusually good for a department-store bike, but very basic. He came back a few weeks later, and asked me to build him a bike. Usually, I asked that the customer come back after a few weeks, to make sure everything is still properly adjusted. In this case, I did not get the chance. The week he took delivery, he rode down to Mexican border, then turned around and rode up the Pacific coast to the Canadian(!) border. Came back very happy - and told me about his trip.
Point is - I never had an unhappy customer.
Today I am chewing on the problem of how to build a cloud. There are a lot of unknowns, and a lot of guesswork. Clouds are a lot more complex than bicycles. And not.
Point of information - the big vendors are nearly as much at sea as sites trying to deploy clouds. Clouds are new. The hockey-stick growth curve means there are a small number of folk with substantial cloud-expertise, and much larger group with little expertise, scrambling to catch up. (As a customer, in part you do not care - as long as the vendor keeps ahead of your needs. Or appears to be ahead.)
This is a new area with a lot of unknowns, and a lot of guesswork. Not unlike my job building custom bicycles for touring, when in college.