Like so many other Princeton engineers, I had loved COS 126 (the introductory computer science class) so much that I was unsure whether I wanted to switch and become a COS major. Initially, Electrical Engineering was my intended major. I knew that the Princeternship program would allow me to explore the differences and similarities between these two fields, and what better company for this than Juniper Networks, a firm that specializes in networking hardware and software.
At 9:30, I met Hal Stern ’84, the alumnus, and Sreya Basuroy (another freshman) at the Ritz Diner in Livingston. We were really lucky to have Hal as our mentor, as he was not only very knowledgeable about the subjects we discussed, but also really kind and funny. Over breakfast, Hal gave a brief history of Juniper, and also explained how he ended up working for the company. He then showed us the different areas that Juniper makes products for, and how the software and the hardware sides work together.
After breakfast, we drove to Shannon Labs in Florham Park, where AT&T research is housed. There we met with Eleftherios Koutsofios, also a Princeton alum, to talk about what it means to really handle data and perform data visualization. Mr. Koutsofios showed us a tracker of all the different data that AT&T has to gather, and the different ways in which the data can be analyzed. The amount of data being generated has increased exponentially in the past several years, and finding patterns now requires breaking up the data into many different pieces, and analyzing those individually. Otherwise, serious problems might only be a small amount of the total data, and would get lost in the traffic.
We then visited Juniper’s office in Bridgewater. Hal showed us several of the hardware products that Juniper makes, and told us about the elaborate process that goes into making such a product. The chip design has to first be created, and then sent out for manufacture. Simultaneously, they begin building the system for the chip, and creating marketing and sales plans for the system that the chip will go into. Overall, it can take up to 3 years for hardware to go from design to shipping. In Bridgewater, we also got to sit in on a conference call, where a startup company pitched an idea to Hal, and Hal discussed how their product can fit in with Juniper’s line. This was really interesting, because we got to see how startups can cooperate with large, established companies so that both benefit.
Overall, I learned a lot and had a lot of fun throughout the day. It was really interesting to see how Juniper is able to work with both software (specifically security software) and hardware (routers, data centers, etc.) separately, and yet encourage collaboration between the two areas. I do not yet know whether I want to do Electrical Engineering or Computer Science, but I do know that whatever I do choose will be important and fun.