Interview: What does a Data Science Product Manager do, and how do you get that job?
Interview with Jack Moore product manager at Qventus
I recently got a chance to talk to with Jack Moore, product manager at Qventus. Qventus is a start up that uses data science to help hospitals to make better decisions. So Jack’s job is mainly around data science products.
So in this interview, he talked about what does a data science product manager do, how he got into it, and he talked about some tips for candidates who are interviewing for similar jobs.
It has two parts:
Part 1: What do you do as a data science product manager?
Part 2: Job Hunting tips as a data science product manager
Q: How did you enter the data science product management field?
My first exposure to data science was at the University of Notre Dame, where I went to school. As an electrical engineer, my senior thesis was using computer vision to distinguish people and cars passing by a video camera .
The process was very interesting, the process of the breath of application of the type of application was fascinating to me. When during my time at Pacific Gas and Electrics, I had the opportunity to develop application which utilize data science. That really caught my eye, really ignited a creative spark in my head and just got me going down the path of what are all the other things that data science might be able to do in the world for the better.
Q: As a data science product manager, what sets you aside from other product managers?
So I think the role of any product manager, or at least one roles of any product managers is to be able to clearly communicate complicated concepts to a wide audience, whether that ’s a software component that you are developing or data science. I like to think of my passion of data science makes me communicating the possibility that data science offers to a wide audience to business folks and other developers things like that.
I also would like to think that my enjoying data science gives me the opportunity to see how it can be applied and the fact that I love it. I do a ton of research and try and learn as much as I can from though leaders of this area how to turn the data science work that we are doing into actionable insights that actual people can use and benefit.
Q: What is a day of yours like at Qventus?
Qventus is a technology company focused on operation software for hospitals, we help hospitals operate more efficiently, a lot of what that involves is like things like detecting things like how long a patient is likely to stay at a hospital, how likely a patient is to readmit within 30 days. So I work on a lot of products that at apply those sorts of algorithms and solutions so that doctors and nurses and the other members of the care team can use the prediction that we are making to actually improve care for the patient.
Q: How you work with other team members?
Ultimately, the hospitals and hospital patients they are our end users, so my job is to learn as much as I can from those user. So I spend a lot of time traveling to hospitals, talking to those folks on the phone. Being on site and learn as much as I can from users, cuz hospitals are incredibly complex. I will never understand as much as a nurse does who spend all their time on the floor, dealing with patients. In terms of engineers, my relationship with them, I like to think of my biggest job is to communicate users’ need to engineers and help them understand why is it that the things we are developing are important, why they have the potentials to create real benefits in that setting.
Q: Why is it important to be able to explain complex data science concept to different stakeholders?
I remember the first time I talk to a PhD data scientist and had them attempted communicate to me how well a model was performing, and we got into a conversation about confusion matrixes and precision recall curves, and area under the curve, I felt that they are speaking a different language. So I think, part of my job is to communicate this complicated concept of how can we tell whether this model or this application of data science is producing the sorts of results that we should expect. That’s a big part of my job, than from there how do the results we are getting translate to actual value. How is a human going to use them what is that mean in terms of likely expected outcomes.
Q: Suggestions to students? Which classes you’d recommend they take?
Every aspiring product manager should endeavor to take at least one programming course during your time at school. I think its an incredible valuable experience, you don’t need to be a good developer to be a product manager, but you do need to understand what it is that your engineers are going through when they are being asked to develop things. Other than that, statistics is a great things to have, understand and quantify the output of a model to put some context around how well it is performing is a very valuable skill.
Q: Suggestions to PMs without tech background?
The biggest thing about product management is being honest about the things you don’t know, and asking your team to explain concept to you. I ask my team all the time, they will use some acronym that I don’t understand or reference some technology that I have never heard of. I will stop them and ask them I don’t know anything that you are talking about, can you explain to me so that I have some context.
My favorite product management book, not sure if it is helpful as much as in terms of trying to bridge a technical understanding, but if you want to be a product manager, I really like “ inspired “ by Marty Cagan. Or “User story mapping” by Jeff Patton. Two of my favorite product books.
Q: Tools you like to use?
Excel: really easy to use, spin up tool to do data analysis.
The excel pivot table might be the most impactful technology, one of the technologies that have the greatest impact in my life. I use it a lot, A lot of time what I use excel or google sheets for is to create rough model. Base on the output we are getting from a data science model, what is the value of it, so if I were trying to predict whether or not a patient is going to readmit to a hospital, what’s the value to the hospital of doing that correctly. What’s the cost doing it incorrectly. By understanding the two outcomes, you can draw a line and say ok this is how the model needs to operating in order to get a net gain in terms of the hospital, and I will use excel for that sort of things.
Q: Data Visualization Tools?
I am very lucky that here at Qventus we have a data analytics team. They help me a lot, they do incredible work to modeling around of data visualization,
My all time favorite tool for data visualization has been tableau. In my past roles, I have interacted with it a lot. It’s a great tool for creating data visualization, and sharing them and making them available to other folks , you can communicate all the things that are going on with your data.
Q: How to earn trust from your engineers?
The biggest thing you can do for your engineers is understand the things that are causing them pain, and represent their best interest. So a great example of this is technical debt . Technical debt is all of the little things that you hack together in order to meet a deadline.Its something that you built when you have ten users, and it is about to break when you have a thousand users.
Its the sort of things where your developers have to program around it, they have to tip-toe around these pieces of technical debt. If you can stand up for your team, and show them that they are priority by saying these concepts that you are telling me are important, are going to be a priority for us. That’s a great way to engender trust between you and your engineering team.
It’s really easy for someone who doesn’t sit with the engineers to say, we have to build this new feature. And we have to fix this critical bug. Do all these things that are very user facing, tech debt isn’t intrinsically user facing a lot of the times. So by saying to the business leaders, I understand what you mean when you need this new feature. But I am telling you if we take this time to solve some of the tech debt we will move a lot faster going forward, We will ultimately deliver more value going forward than if we just kept going in the pace we are going right now.
Q: How to cultivate relationships with other stakeholders?
Dealing with stakeholders can be tough. A lot of times you have slightly different goals. Because of that for myself, the best technique I found is setting a common goal criteria for success. Say we are working together to achieve this outcome. Here is how we know how we can go there. And being able to have that common definition of success, is a really great way to being able to go back later and rebate and say. Just so we all remember, this is what we decided was gonna be success for us. When you are arguing over whether or not some decision is the right one. You can go back and say does this support the definition of success that we are all here to achieve.
Q: Common interview questions for data science product managers?
A lot of common interview questions that I have seen have to do with the mission of the company. People will always ask you why you want to work here. So its always important to find places where you genuinely think you would want to work , and the problem they are trying to solve is worthwhile.
A lot of time you will get screener questions particular to the technology you will be working with, so for data science someone might ask you what’s the difference between precision and recall just so that they know you have some basic understanding.
What’s your favorite product is a really common one, and I know cracking the PM interview is a great book out there that cover stuff like that.
Q: Your favorite interview question?
My favorite question and any product manager or engineer who comes through the doors here at Qventus will hear this from me: I will ask them what is their favorite consumer application of data science or artificial intelligence .
You get a bunch of really interesting answers. You can always tell whether or not someone has a basic understanding and curiosity for data science based on their answer. The great candidates, their eyes lit up and say “ I was just reading about Netflix’ content ranking algorithm or something like that.
Obviously, its important to be genuine, but its always fascinating to hear some of the answers people come up with when you run into a person who is genuinely curious about data science and its applications
Q: Insider tips for candidates applying for Qventus?
One of the biggest thing I love about Qventus how passionate we all are about the outcome we are going to achieve. Everything here is very driven to solve this problem, the inefficiency of the healthcare system. The biggest thing you can take through the door here is a real passion wanting to solve the problem with us cuz they are really hard.
Flora’s words to readers: Thanks for reading my article and maybe watching my video. I didn’t receive any incentives for these content. When I was in school, I had been really lost for a while, because I didn’t know what I should do in the future, and what opportunities that are out there. So I hope with this video series, I can help young people that are just like me to find what they want to be and have ways to achieve it. If you want to support me, you can introduce people you know who are working in the tech industry for me to interview, or you are welcome to subscribe to my channel!
