There and back again…

Hi, in case you don't know me my name is Flavio. This is my first text on medium. I'll try to write a bit about computer/data science research in Brazil. Why? Well, a lot of people already write about this from a US perspective. Topics like academia vs industry, how to get jobs at good places, how to present yourself and your work etc. Not a lot of people write about such topics for the Brazilian reality. I'll try to present my view as a local. Maybe it helps someone, maybe not. The posts from abroad helped me so I'll give it a shot.

Yes, I know I'm writing in English. I'll probably get some criticism for this. I chose English and not Portuguese since it may be able to reach wider audiences. Also, in computer and/or data science English is the language we are often used to use when discussing our work. This brings me to another point I have not yet explained. As you may have noticed, I'm a computer scientist by formation. Nowadays, I also consider myself a data scientist by experience and research. There are many new terms popping up every few months, but I'll stick to these two for now.

So now I believe I can begin my post. I'll talk about industry vs academia in Brazil. To give some context, recently I joined the computer science faculty at the Federal University of Minas Gerais (Universidade Federal de Minas Gerais — UFMG). In Brazil, a lot of public universities will be a UF-Something. U is for university, F is for federal. The other letters usually indicate a geographic location. MG is Minas Gerais, a state, the university is in Belo Horizonte (the capital). We also have state universities, private ones, and federal institutes that I wont discuss for now. Computer Science at UFMG is one of the top CS departments in the country and continent, also my alma mater for PhD studies. Before coming to UFMG, I spent a year as a researcher at IBM Research in Rio de Janeiro. I'll try to share some insights on what made me change as to discuss academia and industry in Brazil.

Maybe if I were joining a top US or some European university I could just begin with, "leaving Company Y to join University X". Most people on the field would understand the benefits and problems of both. Four paragraphs in the text and I'm still on background knowledge. This background is important because we read so much about the US or Europe perspective and a not a lot is said about Brazil or Latin America. So I'll continue. IBM Research opened in Brazil around 6 years ago. You can read about on their website. At the time the news was quite exciting, at least for me, since there aren't a lot of industry research labs in Brazil. Oh, we do have developer or researcher positions from other companies like Google/Microsoft/Startups in the country, but not necessarily a research lab with some space for more academic like research. I viewed (and still do) the lab opening as good news at the time since most PhD students in Brazil end up in another state/federal institution. Usually somewhere close to home. It's cultural in a sense I believe, but it's also due to lack of opportunities elsewhere in the country. As you would expect, a whole bunch of people also go abroad and/or on to less research focused positions. Some good friends followed this path.

From the above you can understand that industry research positions are rare. For a while, academic positions were plenty due to a government program called REUNI. Basically, new federal universities were popping up in large quantities and in a lot of places across the country. New university leads to new jobs. Older ones, like UFMG, also saw a large increase in students and, less so but also significant, in faculty numbers. This is one of the reasons most people went on to federal institutions after the PhD, positions were close to home and there was a whole bunch of them.

I viewed this bloom in university positions as a little bit sad and fateful, I confess. Why? Well, we research computer science and industry is a large part of the game. There is experience and problems that can only be gained in industry. This is all much discussed territory, but again not from a Brazil/LA perspective. Here, there was a lack of such positions for a while. Most of these positions, even if funded by industry, were inside the universities. So you can maybe relate to the sadness. It's a sense that, not matter what you do if you stay in the country you know where you will work. Given that university jobs are very stable in the country, your first job can easily become your life's work. I decided not to do follow on that path initially, but fate, I guess, had other plans.

I did my undergrad at UFCG (Federal University of Campina Grande), also a top institution in CS in the country. Went on the UFMG for MsC/PhD. After it, I was done with academia and went on to IBM Research. Now, with all of this background, we get back to the basics: Why did I return to academia? In a sense, a large part of the discussion is not different from other geographic locations, some points are:

Industry Pros: work on practical problems; higher pay (though cost of living may balance this); hands on experience; no worries about funding; group work is a must (and this is positive).

Academia Pros: freedom to pick your own research; students; learning new fields is easier (comes with freedom/students); you impact people (the students) directly; stability (that can also be viewed as as con).

Some research labs can offer academic benefits and create a almost 100% academic research lab, Google's Deepmind and Facebook's FAIR are probably some examples of them. In any case, I believe is more of an exception than the norm. But there are places like this out there in the world.

Industry jobs were rare and I had one I wanted. Academic jobs were plentiful for a while I did not want one. I was happy with my job at IBM Research. In any case after a while I did find myself looking back to academia in any case. The situation in Brazil also gave me a push.

One of the major reasons I looked were the students. You can also argue that industry impacts people directly, but I'm looking at a one on one level here. Working directly with students is great and should not be underestimated. Also, indirectly a lot comes from academia. If one professor convinces Sergey Brin and Larry Page to look into node ranking on graphs, or more realistically, teaches two good industry developers, his direct impact may be small, but after two hops his impact is much larger. The second is freedom to pick your own research. New areas are blooming and a little bit of more freedom may help in keeping up. I'm being realistic here, it may help, I know that the professor job may take so much time this does not happen. We shall see. In any case, I want to get the discussion back to the Brazilian level.

In comes 2016, oh fateful year. 2016 hit the entire world in full force. You probably know what happened. In Brazil we had an impeachment (or a soft/parliamentary coup as it can also be interpreted) process.

Since 2013, roughly, Brazil was going through an economic crisis. Largest one ever. Some, including myself, and even the government, denied it for a while. I'm not an economist, but by late 2014 (after the world cup and also the election period) the crisis was more clear. People losing jobs and a recession, as you would expect in a crisis. Academic jobs were already becoming rarer at this time. The name Federal on the universities comes with a: we don't charge tuition but we depend on government money. The crisis was affecting the universities, but they are still doing their thing as they always done in Brazil (even as the crisis worsened). Universities survived dictatorships and austerity before, tough times are here again.

I finished my PhD in mid 2015. Thanks to those CS Diplomas and education, I got my job at IBM Research. I also had offers elsewhere, went with IBM due to all of the reasons above. I was focusing on a research career. Economic crisis led to the government crisis. By mid 2016 situation of the country was full scale non-understandable. At least not by me, maybe by the few more enlightened or the many more extreme, but in any case you can find good arguments in favor or against a lot of things that are happening around here. In comes the new Prez, called Temer, that rose to power after the impeachment (he was the vice). Temer began his government with a solution to the crisis, all hail our savior.

Temer's solution is simple, cut public spending. He has a few measures he needs to pass by congress and senate. One of them, called PEC/55 freezes the amount of money the government can spend. Given that most of the best universities are federal, well, money-wise things aren't looking great. Some argue this will end the federal universities, some say it's a great thing that will make better use of the money and cut down on the non necessary. I'm slightly more pessimistic on this. However, I'm not as pessimistic as the more extreme individuals. I believe the universities will endure as I said. This pessimism helped on my decision to return to one as I now explain.

Does it make sense? No, I guess not. Universities are facing tough times and I'm returning to one. Why not stay in industry? Honestly, I was already looking back to academia as I have said. Also, all of my higher education came from the Brazilian UF's. Maybe I'm idealistic, but I'll use my idealism while my few remaining years of relative young age (academic wise) allows. Federal universities are some of the best in the country. A part of me just wants to help out in this time of need. If 2016 is the beginning of the end, well, maybe I can help out a couple of students while the universities do not charge tuition and provide their awesome, relatively free, education. Why not just move abroad? Maybe it will happen over time, but again, I think now is the time understand and debate Brazilian situation. In any case, I'm grateful for my small time but great time at IBM Research.

So these are my professional and personal reasons so far. Some things are still missing from this story. The timeline was not as clear as I presented. For instance, the budget freeze was only an idea when I applied to the faculty position. Moreover, to join the faculty I went through a selection process as you would expect. This process opened up at a good time. I'll try to write more about how faculty is chosen in Brazilian universities later. It may serve as motivation to continue writing. I will also try to compare Brazil's research with other countries. As a final note, it is important to mention that I had personal reasons to return to Belo Horizonte. Bye bye Rio, do get better but never lose your charm.

What is the ideal future? Well, that other industry labs open up in the country. We need innovation and CS research in the country. Brazil has offices from a lot of companies, labs are more rare. If I were to think on ideal terms, labs that come from Brazilian companies would be great. Listen up Santa. Great students are graduating without many options, they will probably emigrate. Universities also need find their way out of the current predicament. A lot is going on: strikes, students occupying campuses, and government austerity measures. It's a time of debates and thinking. I hope people debate and do not kill one or another, as has already happened. Economically speaking, things aren't looking great for my wishes.

That's it for my initial post. It was rather long, I needed to cover a lot of background. Personally, I prefer shorter texts. I'll try to keep it short on the next one.