TrueMotion’s data science team develops the machine learning/AI platform that powers our driving data solutions. It builds the algorithms that help keep drivers safe on the road. What’s it like working on this team of superheroes?
To find out, we caught up with Yuting Qi, our VP of Data Science.
What technologies does your team use?
We use machine learning/deep learning and artificial intelligence, along with sensor physics, signal processing, and information fusion technologies. We use a combination of sensors like GPS and inertial measurement unit (IMU) sensors like accelerometer, magnetometer, gyroscope, and barometer. We develop models on time series data and contextual information, like road information, traffic information and speed limits, etc.
What approach do you take to building our machine learning/AI models?
The whole idea of AI is to mimic human intelligence, so before you step into any data or models, first you need to make sure that you understand the problem itself and the physics behind the problem. Second, you need to think about how you — as a thoughtful person — would predict the outcome of a problem. Then, you need to identify what information you need to make a conclusion. We look at our data and extract information to build machine-learning models. It’s all about simplifying complex problems into solvable ones and figuring out what data will help us crack the code.
What’s the most exciting project you’ve worked on at TrueMotion?
It’s all very exciting! We work with so many data sources. Our data is collected from the smartphone and touches all of the mobile sensors. The more data sources I explore, the more knowledge I gain.
What’s the culture of TrueMotion’s data science team?
We are a very intelligent, collaborative, humble, and supportive team. There are always brilliant ideas bouncing around, we are constantly learning, and we are all open to helping each other — even if we are working on different projects.
Is there any time for team bonding?
Of course! Every week we host data science lunch talks. Team members are encouraged to read a paper, explore new technologies and share what they learn with the team. It’s volunteer-based, so the presenters rotate. We also have brainstorming sessions where we come up with interesting project ideas or new product features that aren’t on the company’s roadmap.
What do you like most about working on TrueMotion’s data science team?
My entire career has been focused on machine learning/AI. The data we work with at TrueMotion is so powerful. It’s so exciting to dig into all of the mobile device’s IMU sensors and to see how new information benefits our products.
What excites you about the future of TrueMotion?
Great question! When it comes to technology, the insurance industry is very behind. Our technology is helping our insurance partners innovate and transform an entire industry.
What’s the interview process like?
Our interviews are hands-on to give you a sense of what it’s like to work on our team. We’ll give you a small project to work on and then you’ll meet with our team. We’ll ask you to present your work to us so we can see how you put a problem into action. We’re interested in your thought process, how you handle data, and how you interpret results. It’s hard to know everything, so we look for someone who’s willing to learn, challenge themselves, try out new ideas, and execute on them.
What’s the best office perk?
All of the company events! Yankee Swap is my favorite. Last year, someone got a live lobster! They ran downstairs to set it free in the harbor.
Time-travel or flight?
Time-travel, so I can see the future.
iOS or Android?
iOS all the way! I only use Android phones when I’m testing our models.
You can bring one innovator to dinner. Who do you bring?
Elon Musk! I love his ambitious visions and perseverance!
Share this post
We’re excited to announce that we’ve joined Guidewire PartnerConnect as a Solution partner to launch the industry’s first claims accelerator…
February 9, 2021