I’ll be honest, I’m a really picky person when it comes to deciding where I want to grow my career. That being said, I was extremely fortunate to come across Geotab 6 months ago. Every day, I get to work with a genuine group of data junkies, work on really interesting projects, and learn from an incredible mentor, who also happens to be my boss!
To put the cherry on top, I got the opportunity to collaborate on an article with Mike Branch, the Vice President of Data & Analytics at Geotab. Mike took this opportunity up without hesitation and wrote very thoughtful responses that I think everyone can learn from.
So stick around if you want to learn more about Geotab, what makes the largest telematics company in the world so successful, and invaluable career Advice.
With that said, I introduce to you Mike Branch, the Vice President of Data & Analytics at Geotab!
About You
1. At what point in your life did you realize that you wanted to get into data science/data analytics and why?
I started off my career as a computer engineer and software developer running my own software development business for just over 10 years. I always had a fascination and passion for creating software that was visually appealing, and in doing, so part of the battle was always how to best convey data to our customers in a way that made sense and was actionable to them.
It wasn’t until I sold the company and moved to Geotab that I was fully immersed in the field of Data Science and I immediately loved it. This was just over 5 years ago and everything I had been doing prior was very analytic driven and not based on machine learning concepts or anything remotely related to AI.
I was immediately drawn to the world of machine learning because it shifted my paradigm of thinking (because it is so different from how one thinks about traditional software development) and it opened up a new world of opportunities to provide next level insight in a more streamlined way. Now I could apply my love for creating visually appealing software compounded with powerful ML-driven insight.
2. What is your favorite concept in data science and why?
The field of reinforcement learning (RL) is very interesting to me because there is a lot of untapped potential here and it doesn’t revolve around having a tonne of labelled data which is necessary for traditional supervised learning approaches. We have largely seen it applied to games and robotics in the past but as cloud-based RL platforms start to further democratize this area by providing requisite RL environments this becomes increasingly exciting for the industry beyond these niche use cases.
3. I feel like this pandemic has given us a lot of extra time. What do you like to do during your free time?
I’m not sure I agree with the "more time" that this pandemic has afforded me 🙂 My wife and I had our third little one (a baby girl) right at the onset of this pandemic on March 15. I still remember going into the Starbucks across the street from the hospital and seeing all of the seating insight pushed to the side coupled with COVID-19 warning signs. It’s our norm today, but was mind-blowing back at the time. So, my pandemic life has revolved around a lot of family time with my two sons and new daughter (all 5 years or younger 🙂 My wife deserves a medal here!
About Geotab
3. For those who don’t really know what telematics is, how would you explain it to them and how does it tie to data science?
Very broadly, the field of telematics captures vehicle movement and health data and transmits that back to a cloud-based environment for ingestion into a software platform.
We deal in the world of commercial telematics for fleets that range in size from 1 to over 100K. We capture data from over 2 million commercial vehicles across the globe and are processing over 40 billion records per day. Our 80-person data science team then generates insight from this data to help our customers. We answer questions like: Which vehicle is best for me from a fuel economy perspective? When will the battery in my vehicle fail? Which of my vehicles are being underutilized? Which vehicles can I transition to electric vehicles?
Geotab has the richest vehicle dataset in the world to provide value to our customers from location data to massive amounts of engine diagnostics data including RPM, windshield wiper activation, EV charge state. Working with this volume and variety of data is a huge draw for data scientists looking to make a difference and as one of my colleagues says → "They come for the data, but stay for the culture."
4. ABI Research ranked Geotab first last year – what makes Geotab unique from a data analytics perspective that drove it to achieve the #1 telematics company in the world?
We are a data-driven culture at the very core. That message is clear from our corporate tagline of "management by measurement" and is driven from our CEO who personifies what it means to be data-driven in how he runs the organization and the culture he has created.
Where I think we differ is in our approach to data problems – we focus on value-creation for our customers and not on monetization. It may seem like a small nuance, but if you are hyper-focused on data monetization it can lead you to create products that aren’t in the long-term interests of your customers. We also have a strong belief in leveraging data for good and have made available numerous aggregate datasets on our open data portal Geotab Ignition for developers and data scientists to use to create applications that lead to safer driving and safer cities – everything from hazardous driving areas in cities to hyper local precipitation (drawn from windshield wiper activation). Our team also puts a big emphasis on rapid experimentation and creating data-enabled MVPs (minimum viable products) which really facilitates the incubation of new ideas and furthers Geotab’s position as a market leader.
Advice to share
5. What do you think is the most important data science concept and why?
Treating data as a product. And I don’t just mean data cleaning. Oftentimes in organizations, data is treated as a byproduct of another process and what people must realize is that data is the fuel that powers AI and for us to have good AI and make good predictions, we must have good data. This starts with treating data as a product in your organization. By this I mean data catalogs, data lineage, data discoverability, usage policies, anomaly detection, monitoring, residency, ingestion pipelines, privacy and more. There is a tonne of work that is sometimes less than glamorous but absolutely essential to having a sustainable data practice that produces good models and drives how even better insight.
6. What skill do you think people overlook, but should be spending more time on in the data industry?
SQL. It may seem relatively simple, but data scientists (in my opinion) generally come with very good chops in Python and R, but with mediocre SQL experience. And by SQL, I don’t simply mean the ability to write a select statement – I mean complex joins, windowing, geo-spatial operations, handling massive amounts of data in efficient queries, etc. With platforms like Google BigQuery and Snowflake, more and more we are seeing how much can actually be done to prepare and pre-process your data just leveraging SQL. In fact, in many cases you can leverage advanced SQL syntax to create and train basic models (many of which can provide answers to 90% of your business needs).
7. Lastly, do you have any advice for those who are just starting their career in data science/data analytics?
In my opinion, I believe those that infuse software engineering practices into their data science career are in the best position because they know how to take models from inception to production and they know what it takes to scale out a model across an ecosystem. Not only that, but these people have the ability to wrap a piece of software or API around their newly-developed model and can quickly highlight the benefits to end-users and achieve adoption more quickly. Otherwise, models run the risk of staying in the lab and not seeing the light of day.
Thanks for Reading!
I hope that you found this article insightful and enjoyable! Thanks again Mike for sharing your experiences and wisdom to the rest of us :D.
If this has sparked your interest in Geotab, we are growing really fast and I urge you to check us out. I’m also more than happy to connect with you on LinkedIn if you have any questions!
Not sure what to read next? I’ve picked another article for you:
Chatting with a Data Expert From Google, Christina Stathopoulos!
and another one:
A Complete 52 Week Curriculum to Become a Data Scientist in 2021
Terence Shin
- If you enjoyed this, follow me on Medium for more
- Interested in collaborating? Let’s connect on LinkedIn
- Sign up for my email list here!