Starting September, 1997, I've been permanently employed at startup Covad Communications. At Covad I'm a network operations engineer, building and supporting an ATM and frame relay network. This is a first for me, the closest to the front lines I've ever been. My day spans hacking Excel spreadsheets, developing and writing troubleshooting processes, designing ATM and FR connections, developing methods to verify connection integrity, unbelievably time-consuming equipment-ordering, and assorted firefighting.
One of the nicest things about working for Covad is the 1.1 megabit ADSL line into my house. Whee! Dialup? Feh!
Before Covad, I was in a black hole, from September 1996 - August 1997, working a very interesting contract for AT&T. To do that contract, I took a temporary (turned permanent) leave from my all-time favorite job, at the Numerical Aerodynamic Simulation Facility (NAS) at the NASA/Ames Research Center.
Lots of the stuff I did is still listed in the R & D part of NAS' most excellent LAN group, in particular the NAS Local ATM (which is unfortunately missing some links to graphics, but the content is still there).
My job at NASA marked an important career change, from software engineering to network engineering. I gladly traded bug meetings, Makefile hacking and religious arguments about source control for the joy of seeing ping ping.
The foundation for the shift into networking happened during my stint as a graduate student at UCLA, where I struggled through as much CS theory and math I could handle. This sharpened my mind, helped developed logical reasoning skills, and helped me develop a structured approach to problem-solving. I picked up a fact or two about networking also, and made some great friends, but most importantly, the experience helped organize my muddled mind. Not being a natural student, and having almost no talent for computer science, I am very proud that I survived and earned my MSCS.
The AT&T gig marked a small shift in my networking career, from local area networking (ATM and HIPPI, primarily), to wide-area networking (ATM again, and now, frame relay). Going from LANs to WANs means trading a Unix workstation for a Win95 laptop, and spending all your time on the phone instead of crawling around under raised floors in freezing computer rooms.
I love being a network engineer. After college, I thought I'd stay technical for 10 years, then go into management. Heck, no! 12 years later, I have no intention whatsoever of getting away from nitty gritty techie stuff. Besides, I'd stink at management.