Luca Cazzanti says his wife calls him an “International Man of Science.” The “international” part comes from having moved back and forth between Italy and the United States a number of times, plus his many travels and his job at NATO. The “science” part comes from working in research and development, both in academia and industry.
We’re going to just go ahead and call Cazzanti GeekWire’s latest Geek of the Week.
As a machine learning and data scientist, Cazzanti has worked in diverse areas: maritime traffic analysis and prediction, consumer behavior analytics, digital music identification and recommendation, fraud prevention, digital telecommunications, and underwater acoustic signal processing to name a few.
“I suppose variety characterizes my professional journey, always in contexts where rigorous science and data analysis are essential,” Cazzanti said.
Born and raised in Milan, Italy, until the age of 14, Cazzanti came to the U.S. when his father’s job with IBM brought his family to Minnesota. Assignments at IBM for Cazzanti’s dad caused the family to move between the two countries three times. Cazzanti eventually finished his master’s degree in electrical engineering in 1998 and moved to Seattle, where his first job was developing algorithms based on neural networks and radio-frequency fingerprinting to detect cloned cell phones.
After jobs at MathSoft/StatSci and the Bellevue, Wash., startup Cantametrix, Cazzanti landed at the Applied Physics Laboratory at the University of Washington, where he stayed for 10 years. He worked on Department of Defense-funded applied research projects in signal processing, pattern recognition, wireless communications, and underwater acoustics. Cazzanti also decided to pursue his Ph.D. and called working full time and going to school “grueling.”
“I stayed at the APL-UW for a while after my Ph.D. then decided to return to the private sector, because I wanted to get into that new thing called big data,” Cazzanti said. “I was lucky to join Globys, a Seattle-based scientific marketing company, in their data science group, which is now their Amplero spin-off.”
Cazzanti said the experience helped him grow into a data science professional very quickly, and his next pursuit was an opportunity to join NATO in Italy and bring his expertise to bear on maritime vessel traffic analytics.
“I really wanted my wife and kids to experience Italy from the inside, not only as tourists. My kids are now seamlessly bilingual and very adaptable, and have developed the ability to spot similarities and differences between the way people in the U.S. and Italy get through life.”
Now Cazzanti is coming back to the city he considers home, and he’s in talks with a Seattle startup about what he calls a tech leadership role.
“My wife has missed working in Seattle but has also enjoyed living in a different reality and trying on a different lifestyle for a while,” Cazzanti said. “Scientist posts at NATO are rotational, and the end of my term is approaching. We are heading back to our home in Seattle thankful for and enriched by this international experience, ready for the next stage. Seattle has changed, and so have we. I can’t wait to find out what happens.”
Learn more about GeekWire’s Geek of the Week, Luca Cazzanti:
What do you do, and why do you do it? “As an applied machine learning scientist I am always straddling two fields: the field of machine learning and the application domain to which I am applying it. In all the roles I’ve had, I’ve always connected these two facets, so you could say that what I do is bring machine learning concepts and tools to bear on a particular subject matter. I very much enjoy helping experts from different domains learn from their data and unleash their creativity through data-driven information products.
“In my current role at the NATO Centre for Maritime Research and Experimentation, I develop analytics platforms in support of maritime domain awareness (MDA). MDA is the understanding of economic, environmental, security, and safety factors connected to the maritime domain (seas, ports, vessels, infrastructure, etc.). For example, one aspect of MDA is tracking vessels in real time to ensure safe passage, monitor suspicious activities, enforce fishing regulations, or optimize port operations and logistics. Another aspect is measuring and predicting broad economic indicators, like the volume of traded goods, number of vessels traveling between countries, which routes vessels follow and whether they are carrying oil or bananas, or fishing.
“Currently MDA analysts carry out many steps manually, spending time collating data from different sources or setting up ad hoc vessel-tracking rules. This is a drag on their daily workflows, and limits the achievable sophistication and granularity of their analyses. I work on computational approaches to MDA that include machine learning, analytics at scale, and effective visualization. There are two main goals. One is to streamline the most rote tasks, so that analysts can focus on insights, policy, and strategy rather than on pulling data into a spreadsheet. The other goal is to provide thought leadership through more sophisticated analytics that help MDA stakeholders achieve new workflows that they had not even imagined possible. A way to achieve these goals is through analytics platforms that automatically learn and predict the patterns-of-life of maritime traffic and serve the discovered insights to the MDA analysts through interactive interfaces.”
What’s the single most important thing people should know about your field? “Most people do not know that 80 percent of worldwide trade happens by sea. As techies we more often look at drones, 2-hour shipping, on-demand, transportation services, and (coming soon) autonomous vehicles as examples of positive disruption. These innovations are impressive, yet they operate on land, in the last mile of the supply chain. At the other end, maritime shipping is the backbone of worldwide commerce: your new automobile and the oil that powers it, the grains for your multi-grain artisanal bread, organic bananas, etc. arrive by sea. It’s a huge industry with many different stakeholders that include ship owners, shippers, insurers, commodity traders, hedge funds, port authorities, the military, law enforcement, safety and environmental agencies, and international organizations like NATO and the UN. There is an opportunity to develop a data-driven, self-service analytics culture in the maritime space, powered by information products based on machine learning and modern data analytics stacks. They could ease the pain points of the various stakeholders and radically change the maritime industry.”
Where do you find your inspiration? “My two wonderful children inspire me to be a better person and help me keep things in perspective every day. Everything follows from that. In addition, I find inspiration from knowing that what I do is helpful to others and positively impacts their everyday activities.”
What’s the one piece of technology you couldn’t live without, and why? “I would not be able to do my job effectively, or even get through the logistics of daily life without computers, the internet, cell phones, and reliable transportation. Those of us who are privileged with access to these resources have come to rely on them and take them for granted, but on occasion I have been able to do without them. I’ll say this: a reliable espresso machine, that is something I can’t give up.”
What’s your workspace like, and why does it work for you? “At work I have the biggest private office I’ve ever had, and probably ever will. But the real perk is the beautiful location of our building, right on the water, overlooking the Gulf of La Spezia, across the bay from beautiful villages perched on the hillsides.”
Your best tip or trick for managing everyday work and life. (Help us out, we need it.) “I find that exercising helps me think through tough problems. I try to tackle my toughest challenges, which require focused thinking, in the morning. Then I take some time to exercise around lunch, and reserve meetings for the afternoon. I found that exercising helps my brain work in the background in a more diffuse mode of thinking which helps me be creative. After my afternoon meetings I often get a second wind and am able to make more progress on my morning’s challenges.”
Mac, Windows or Linux? “I am multi-denominational, by accident, not by choice. I have a MacBook Pro and a Lenovo laptop running Windows. My wife and I own an iPhone, a Nexus 5, and Nexus 7 tablets. I do most of my computing on Linux VMs, mainly on the Python/Jupyter stack.”
Kirk, Picard, or Janeway? “I am not a Trekkie, but I remember watching the original ‘Star Trek’ series dubbed in Italian while growing up in Italy, so I am going with ‘Capitano Keerk.'”
Transporter, Time Machine or Cloak of Invisibility? “A transporter to move back and forth between the U.S. and Italy more quickly and frequently, sans the jet lag.”
If someone gave me $1 million to launch a startup, I would … “build the go-to maritime analytics platform based on machine learning to bring a data-driven, self-service analytics culture to maritime commerce and disrupt the maritime industry.”
I once waited in line for … “Noam Chomsky. It was 20 degrees below zero outside, people rushed into the auditorium to stay warm and overwhelmed the ushers … short story is I ended up watching his talk on closed-circuit TV while sitting on the floor of the beer garden in the student union, even though I had waited in line for a proper seat for a long time. Fun times!”
Your role models: “My parents. They were born during WWII in Italy and grew up during the Marshall Plan. They basically started from nothing and built a solid, safe middle-class life for me and my brother through quiet persistence. Their ambition was simply to give me and my brother opportunities and choices in life, and it is in large part thanks to them that I am in such a happy place personally and professionally right now. I strive to do the same for my own children, but I recognize that I am starting from a position of great advantage that my parents did not enjoy in their time.”
Greatest game in history: “When I think of games, I think of soccer matches. My dad’s generation would say Italy vs. Germany 4-3, semifinal of the 1970 world cup. But I say Italy vs. Germany 2-0, semifinal of the 2006 world cup: intense, both mental and physical, tactical, two goals in the final minute of a wide-open overtime, and the underdog beat the home team (and eventually the World Cup). Solid!”
Best gadget ever: “Corkscrew.”
First Computer: A Spectrum ZX. As a kid, I used to copy BASIC source code from computer magazines into the Spectrum, save the code to cassette tape, and then load the code back from tape to play games. My brother and I played JetPac a lot, a space invader-type of game. He found a bug in the game where if you left your player in a specific position on the screen, it would never get hit by enemy fire. He got lots of high scores in this way … until I found out.
Current phone: “An iPhone that’s too old for disclosure on a respectable tech web site like GeekWire.”
Favorite app: “Put me in front of a Jupyter/IPython notebook and I’m a happy man.”
Most important technology of 2016: “Voting from abroad by email. Thank you King County!”
Most important technology of 2018: “Pervasive IoT driven by machine learning/AI.”
Final words of advice for your fellow geeks: “Developing a tech product is more about people than technology. If your product is driven by data and based on machine learning, the easier part is throwing algorithms and tools at the data until something sticks. The much harder part is listening to the stakeholders, understanding their pain points, and honing in on the fundamental issue with which they are struggling, so that you can focus the tech on solving it. Throughout this process you need to keep the stakeholders engaged as partners, and educate them on the benefits of the innovation you are proposing, but always keeping your ears and eyes open to their unspoken, unmet needs. It takes patience, acceptance, and respect for the uncertain space this endeavor occupies. Enjoy the ride!”
Website: Luca Cazzanti
Twitter: @lucacazzanti
LinkedIn: Luca Cazzanti
