Of A Very Shiny Crystal Ball Called Google (Part 2)


Zürichsee, from Arboretum, a small park at the very top of the lake

Exactly one day after I published this blog post here, a recruiter from Google contacted me, saying that there is an open position fitting my profile:

Analytical Linguist, NLU in Machine Intelligence (Thai and Indonesian)

Understanding natural language is at the core of Google’s technologies. The Natural Language Understanding (NLU) team in Google Research guides, builds, and innovates methodologies around semantic analysis and representation, syntactic parsing and realization, morphology and lexicon development. Our work directly impacts Conversational Search in Google Now, the Knowledge Graph, and Google Translate, as well as other Machine Intelligence research.

As an Analytical Linguist, you will collaborate with Researchers and Engineers in NLU/Machine Intelligence to achieve high quality data that improves our ability to understand and generate natural language systems. To this end, you will also be managing a team of junior linguists and vendors to derive linguistic databases as well as propose direction for approaches to language specific problems.

Target languages: Thai and Indonesian

My first reaction was… amazed. How come, just right after I posted a writing about my epic failure at the previous interview?? 😀 What amazed me even more was that this opportunity answered my question back then, quoted exactly as it is: “what else could be the reason to work at Google other than having something ‘shiny’ in your CV? :p”

Answer: I could work on natural language processing (NLP), which I really love doing, for Indonesian language, my mother tongue! Well, of course I can always work on NLP for Indonesian language if I really want to, but working on something that could have a very big and direct impact to the society? This is my chance! Probably a chance that comes once in a lifetime…

There were two things that must be considered before I could continue. First, this is a full position, not an internship. Second, the position requires a relocation to… Tokyo, Japan. Since my PhD study is not yet over *although almost, I hope*, getting this job means juggling two responsibilities, which I can imagine will be very tough. Relocation to Japan is actually quite compelling, I’ve always wanted to visit Japan since I was still in high school, even though living in Tokyo seems a bit stressful *compared with living in a small and peaceful town in Italy :p*. Anyway, in the end I decided to take the chance, because… who wouldn’t? 😉

I went through 7 interviews in total… phew.

Actually 8 if we include the very first phone call with the recruiter. The first one was a phone interview, and unlike for the software engineering internship position, there was no programming question involved. The next five interviews were onsite interviews, at Google Zurich, although all of them were Hangout interviews, not meeting the interviewers in person. The last interview was exactly a month later. It was again a Hangout interview, and seemed to be the decisive one, probably because they still couldn’t decide based on the previous interviews.


Before the interview, having lunch at Arboretum, with a very nice sunshine and a refreshing view of Zürichsee

The onsite interviews were really fun, exhausting *starting at 3 pm, 4 hours of back-to-back interviews in total* but fun. Most of the questions were about Indonesian language, what are its characteristics that can be difficult for natural language processing. No programming questions whatsoever, but there were several open questions about managing an annotation project.

During the short break in between, a short tour around the office was given. It’s a cool place, as how Google offices are always advertised and talked about. The view from the rooftop is especially nice, Zurich is indeed beautiful.

After all of that, I really hope I could write a good news here. Unfortunately no, I don’t get the job :).

“As to the rationale, the team decided to move forward with another candidate they felt had more experience dealing with large scale linguistic-related projects.”

I have to say, I like the idea that perhaps I managed to get to the final round, but in the end lost to someone who is better than me ;). That wishful thinking helped me recover from the ‘heartbreak’. Along with some thoughts like “living in Tokyo must be horrible, too many people, too expensive, will need to learn a new language with different characters”, or “good for me, now I can peacefully finish my PhD without any distraction.”

However, I cannot shake off the feeling that I know the person who gets the job :D. I mean, a PhD student (or has already graduated from PhD) working on NLP, and a native Indonesian language speaker… there are only a handful of people with those qualifications that I usually meet at an NLP conference. For example… this, this, this, this and this *dear companies/universities, if you ever need NLP researchers who are also native Indonesian speakers, there! :D*

Haha, yeah.. during the preparation for the interviews, I did some research not only on the advancements of NLP for Indonesian languages, but also on my possible competitors. As Sun Tzu said, “If you know the enemy and know yourself, you need not fear the result of a hundred battles.”

Let’s hope that there will be a next part… Part 3, Part 4, Part 5, etc. I don’t mind… I will always try my best if given a chance ;).

Leave a Reply

Your email address will not be published. Required fields are marked *