Career Path: Ghinwa Choueiter, Data Science Lead at Sense
What does the career path and day in the life look for the Data Science Team Lead at Sense?
We interviewed Ghinwa Choueiter to find out!
Where did you grow up? What did you parents do for work? What was your very first job?
I grew up in Lebanon in a suburb of Beirut. My mother was a mathematics and science teacher and she decided to stay at home after having children. My dad, who is now retired, was an anesthesiologist. I credit them both for my love of math, engineering, and problem-solving.
Growing up in Lebanon, first during the civil war and then during a struggling economy, it was pretty hard for a teenager to get a summer job as is the custom here in the US. It was not until my undergrad years that I landed my first job fixing PCs in a computer shop in Beirut. Not sure if it counts as a job since I didn’t get paid, but I definitely enjoyed hanging out with the IT team, pulling computers apart and fixing them.
Can you talk about your time as a graduate student at MIT, where you studied Electrical Engineering and Computer Science? What types of research projects did you work on? Is there anything else that you’d like to highlight from the years there?
I joined MIT as a graduate student in 2002. I had a fellowship my first year so I explored a few options as I figured out what I wanted to do my PhD research on. I took a hardware class on low-power analog electronics where I designed biologically-inspired circuits. I also took a class on speech recognition where I learned about state-of-the-art tools to build systems that could recognize and understand human speech. I enjoyed circuit design, but I found ASR (automatic speech recognition) exhilarating. I decided to learn more about sound production, acoustic and language models, hidden-Markov models - all the fields that made ASR so interesting, and I joined the Spoken Language Systems (SLS) Group in CSAIL, MIT. While at SLS, I worked on extracting better acoustic features for ASR using wavelets. I also tackled the out-of-vocabulary problem: ASR systems typically know a limited vocabulary (could be a million words but still limited!). That means if a user says a new word to an ASR system (e.g., Ghinwa Choueiter!), it will definitely misrecognize it and will often mess up surrounding words as well. I worked on improving user experience by handling out-of-vocabulary words such that the system knew the word was unknown and did not mess up neighboring ones.
After obtaining your PhD from MIT, you started your career as a speech scientist at a startup called Vlingo (a personal assistant application that was popular for Android). How did you get connected to this company and how did you know that this career path was even an option?
I wrapped up my thesis at the end of 2008 when the job market was suffering. I recall considering jobs in the financial field and discussing it with my advisor, Jim Glass. His words to me where: “Why don’t you stick a dagger in my heart!” Jim then mentioned a couple of companies in the area doing speech recognition and one of them was a startup called Vlingo. At the time, I was hesitant. Startups are risky and my Lebanese upbringing had taught me that you get a job and hang on to it for the rest of your life! I did, however, apply to Vlingo because the job looked very interesting and it was a nice match to my graduate studies. I enjoyed my interview there, particularly everyone’s enthusiasm and dedication to the company goals. I decided to take a chance and it was the best career decision of my life. It also changed my opinion on startups in general and I’ve only ever taken jobs with that kind of company. I love the fast-paced environment, the small-company feel, the fact that every decision you make and the problem you solve has a dramatic effect on the company path as a whole. You are very visible in a startup, and I guess my ego needs that.
Since then, you’ve held data scientist and machine learning positions at other tech companies before your current position. If someone was interested in a career in data science and machine learning, what should they do in terms of building the right foundation for their career?
After Vlingo was acquired, I decided to try something new and unrelated to speech recognition. I worked at both DataXu, an Ad-serving company, and Hopper, a travel company. It was around the time I joined DataXu that I became fully aware of Data Science as a field. Don’t get me wrong, I - and many others - had been doing data analytics and machine learning for a while, but didn’t realize the field had a name and I also always considered myself to be a speech scientist. As I transitioned into Data Science here are some things I learned over the years:
Be curious about your data: It’s called data science for a reason! Data is your biggest asset so before you start tackling your problem, explore your data, question it, slice and dice it and get an intuition for the challenge at hand.
Keep on learning: Your workplace is filled with smart people who know a lot of things you don’t know. Whether they are senior or junior, in your team or not, keep an open mind and listen to what they have to say. You will probably learn some valuable lessons.
Work on your communication skills: It doesn’t matter how much you’ve perfected data analytics, if you don’t know how to tell a story with your data then you’re missing an important skill. When you watch a talk that you enjoy and you learn from, ask yourself what made that presentation unique and useful and do learn from that.
Can you share the high-level responsibilities of your current position at Sense and what your team is working on?
I lead the data science team at Sense which entails setting long and short-term goals as well as providing technical guidance for our various projects. We dedicate significant effort to the technical challenge known as load disaggregation. This means, in order for the Sense app to provide useful appliance-level insights to our users, we - the data science team - need to design algorithms that learn the distinctive power signature of different devices. In addition to identifying appliances, we are also developing techniques to detect faulty or misbehaving devices as well as ways to help our users increase their energy awareness and savings by identifying high-consuming devices.
Day in the Life
Coffee, tea, or nothing?
What time do you get into the office?
Somewhere between 8 and 9 AM.
Every day is different, but can you outline what a typical day looks like for you?
Morning: Wake up around 5:30 and exercise before my child wakes up. Prepare breakfast with the family and pack food for the day. Drop off the kid at daycare on alternate days. Get to work. Check my calendar to go over meetings for the day.
Afternoon: Either coding/working on a particular technical project, in a meeting, or organizing notes before or after a meeting.
Evening: Pick up the kid from daycare on alternate days. Prepare and enjoy dinner with family. Work. Read. Sleep.
What time do you head out of the office?
Somewhere between 4:30 and 5:30 PM.
Do you log back in at night or do you shut it down completely?
I do log back in at night after dinner to make up for some of the lost time during drop-off and pick-up.
Any productivity hacks?
I keep a daily status blog of my work and I write it early in the morning after checking my calendar. It helps me plan my day and keeps me on track.
What are the 3 apps that you can’t live without?
I am not much of a phone app person, but I do use Whatsapp to call my family in Lebanon and of course Sense to check my energy usage, whether laundry is done, and to make sure I did not leave any - mainly - heating appliance on.
What professional accomplishment are you proudest of?
Becoming the team lead for the data science at Sense definitely stands out. I was initially asked while I was still on maternity leave. I just recall being very tired and sleep-deprived and wondering why my boss would give me that responsibility. It was a great opportunity though, one that I was looking forward to, so I accepted on the spot. It’s been a great learning experience and I am grateful it is with the Sense Data Science team - the best!
Who do you admire or call upon for professional advice?
I am lucky to be surrounded by many people whom I consider mentors who have influenced my career. I will, however, single out Ryan Houlette, the VP of engineering at Sense. Ryan has had a significant influence on some of my career choices and I lean on him for both technical and management advice.