The first skill required of the first software developers, all self trained, was to take raw data in and output a printed report/product/answer. The term used was data processing. The raw data from the viewpoint of the software was often still laboriously prepared by persons, some of whom were highly skilled. Still people waved lots of money to pay for it.
In 1965 I worked at Control Data Corporation in Houston Texas. Starting as a delivery truck driver, I returned punched card decks to customers along with the printer response to their attempt to write a successful Fortran(mostly scientific) or Cobol(mostly Business) program. As I experienced the face to face encounter handing the delivery to the originator I always paused to get their reaction. Some were satisfied as they turned their back to me and walked back to their office. If they were not satisfied I was there to help them analyze their mistake. The first time this happened I knew nothing about solving a software problem. I read every manual I could find and talked to every mentor I could find. In 1965 there were zero classes available on how to be a software developer.
After about 6 months I was approached by a 20 something manager, Harry Scott, who gave me a chance to take over for a let go programmer, a school teacher who tried her hand at programming. I spent the next 2 years working on completing and debugging several big software systems. The first was the largest, 4 big 3 foot drawers of punched cards, a Fortran language program. The payroll program paid, printed the pay checks, and accounted for credit payments for all the stevedores who loaded and unloaded freighters, ships, in the four port cities of Texas, ~$27,000,000 per month.
There was oil field analysis that plotted the oil deposits logged by Schlumberger, a big oil business in Texas. I was thrown a project to help a large engineering company, Raytheon, to use a new optimization software app to maximize things, whatever.
I believe that this nation should commit itself to achieving the goal, before this decade is out, of landing a man on the Moon and returning him safely to the Earth (President John F. Kennedy).
The most interesting projects revolved around the NASA Space Center and the Mercury and Apollo space programs. Control Data Corporation's Control Data 3600/3800 was the most powerful computer in its day. I have heard that the latest iPhone is more powerful; somewhat like comparing an Apple to an Orange. Anyway first we had to build the Space Center. As we know there was a space race going on with Russia. NASA was doing everything in three parallel paths to insure success. I worked with NASA to enter 3 separate activity schedules into a new software application known as Critical Path Method (CPM). NASA Houston Space Center was finished in record breaking time. During the Mercury and Apollo space launches and missions our Control Data 3800 was booked by NASA to process mission data. NASA Engineers would pore over mission data for months between missions. I was there the night a young girl, a prodigy mathematician, helped rescue Apollo 13 lost in space.
My two years at Control Data Corporation was packed by many hours of work and study. Sometimes I worked 3 days without going home, sleeping on the computer room floor for some program to complete a test run. Somehow I managed to go full time at night to the University of Houston and graduate. In one major class I scored a 96 (A) on final but did not study at all for the mid term and scored 20 something an (F). According to averaging I was to get a D. I made my case to the professor and when I got my grades; I got a C.