Datacast Episode 96: Data Science Training and The Power of Education with Merav Yuravlivker
The 96th episode of Datacast is my conversation with Merav Yuravlivker— the CEO and co-founder of Data Society.
Our wide-ranging conversation touches on her early career as a teacher, the founding story of Data Society, the importance of industry-tailored data science training, her journey bootstrapping the company and changing the business model, engaging women in data science, the evolution of education, and much more.
Please enjoy my conversation with Merav!
Listen to the show on (1) Spotify, (2) Google Podcasts, (3) Apple Podcasts, (4) iHeartRadio, (5) RadioPublic, and (6) TuneIn
Key Takeaways
Here are the highlights from my conversation with Merav:
On Teaching Elementary Special Education
Fresh out of college as a 21-year-old in New York City’s public schools, I wanted to do the best I could for my students. What I found coming into the classroom is that teachers have many hats. Especially at a younger age, you are not just teaching but also an emotional coach for the students. You are bringing that level of support and learning how to juggle that.
Over the first few months, I was still finding my feet. I started to grow into my own as an educator by understanding how to set objectives and share them with students. What I did to help my students get invested in their education was walk them through and say: “What are the goals you have? What would you like to read by the end of the year? Let us work backward — put a milestone in there and work towards your ability to read.” When they have that objective in mind (which they help set themselves), it creates a feeling of autonomy for them and investing in their own education (which is crucial at a younger age). By developing this early-stage data tracking, you are helping these students orient on a goal and helping them work through it.
On Being An Exceptional Teacher
To be an exceptional teacher, you have to have an extraordinary amount of patience. Children are not fully formed adults yet. It is important to remember that their perspective differs from an adult’s. Having a lot of patience is key.
Another important factor, which is difficult for many teachers, is to be able to take care of yourself. Teaching is a grueling job: You are up early and get out late. You do not take breaks. It is not like you can hop out of the classroom to get a coffee like people who work in offices do. It is very demanding, and you have to be able to take a step back from the classroom and find ways to re-energize yourself (whether it is taking a bath, going for a run, going out for drinks, etc.) so that you can be a better teacher when you are in the classroom.
On Her Time At IBO
After a few years in the classroom, I realized that I wanted to be able to affect change on a larger scale. I love teaching very much and want to stay in education. I found an opportunity with the IBO — an organization that crafts and helps deliver an accelerated program for students between kindergarten to high school. I was brought on board to help with a worldwide assessment that students take every year, recruit instructors, and set up online training. All of a sudden, I was bringing in my education background and using that in a bunch of different facets. I traveled and met many instructors and teachers, which shaped how I think about assessment and training today.
I still kept my feet on the ground by teaching GRE part-time. It was a pleasure to work with motivated students and walk them through key strategies they need to know to do well on this exam. For me, this period allowed me to grow in other ways in education while still being able to teach in the classroom.
On Founding Data Society
Back in 2014, I had been in my job for a few years and was thinking about where would be the next stage of my career. At that point, I knew I wanted to do something different but did not know what. At that time, I met my co-founders: Dmitri Adler (who had a lot of experience as a Wall Street analyst) and John Nader (the GC counsel with experience in operations of organization). All three of us independently had worked with data and saw that it was becoming more important. We also saw through our journeys that it was difficult for us to learn how to use data effectively.
There were not a lot of customized materials for me at the time. It was difficult for me to make the connection between how to use data in theory and how to use data in practice. I had taken some statistics in college but was not a math major, and I had never programmed before. So all three of us found difficulty in finding the type of education that we needed and the type that we knew a lot of other professionals needed as well. So we all separately identified this problem and ended up getting together: “If this does not exist, since we all have skill sets that are complementary to each other, let’s start a company and build our first program to help professionals learn these skills quickly and effectively.”
That was back in 2014. In the first few years, we were learning data science. We were learning how to run a business. We were making connections and speaking to students. Initially, we started with individuals. Then we saw many of our students asking us to train their teams. Thus we saw a need on an organizational level to train groups of people so they could communicate well with one another. That is how we have grown into the company we are today. We now deliver full-scale programs on an annual basis to help build a data-driven culture within an organization. At the same time, we built out the solution side of the house because we got these requests as we were teaching.
Our mission is to help people use data better. I genuinely believe that if we can get to a big enough scale, we can shift how industries work and think. All of that is because we know how powerful education is and how important data is. When you put those two together, it can make great things happen.
On Getting Into Programming
It feels a bit crazy to think that I started a data science company without a data science background per se. I had never really programmed before, so I found resources online to start cobbling it together. What was helpful was speaking to other people in the field who were already using these languages because they helped me understand the use cases behind why it is important to know programming. They also gave me references for more resources to learn from.
R was my first language. It honestly felt like the coolest thing in the world because it never felt attainable to me as an education major to do things such as programming, creating data visualizations, and cleaning data quickly to get some insights. I remember feeling so empowered that I could do that.
I fell in love with solving challenges. With the job that I have now, it is about doing things I have never done before and being able to find solutions to challenges to help move our company forward.
On Industry-Tailored Data Science Training
It is not that there is a lack of training resources, but the challenge we saw is that professionals do not have time to spend hours searching through videos. They need to understand the real-world challenges of data, which has to do with data cleaning. Given those two factors, when working with organizations, we talk to them about their use cases and datasets so that we can incorporate that into our content. We provide coding templates they can see step-by-step on how we go through the process. They can take that template and reconfigure it with their own data to get these skills up and running even faster.
It is vital that people can connect with the use case because that is how they are motivated to learn it faster. That is why it is critical to have the type of tailoring and an instructor in the room to answer tough questions professionals face in their careers today: How to clean a particular dataset? How to think through data analysis and present that? Those questions can be answered only by experts and instructors in the classroom. Giving that extra level of support is key to the success of our programs.
We are based in Washington, DC, so about 50% of our work is with federal agencies. That is great because the federal workforce is excited to be trained in data. Many of them have a mandate to become more data literate. It is very much in line with their objectives. On the commercial side, we found that finance, healthcare, and defense contractors are key areas where the expertise to work with data needs to grow very quickly. Especially in healthcare, if we think about the past few years with the COVID-19 pandemic, the ability to use data is a life-or-death situation. How do we distribute vaccines? How do we track where outbreaks happen? How do we sequence new variants that come out? All of that requires data, so being able to understand how to use that effectively is critical. We saw that there were needs in these areas, which is how we have defined our verticals.
On Client Relationships
The biggest challenge across consulting and training is ensuring good communication to establish trust. We are good at what we do, but if we do not communicate that well to the client, they might not understand all the work that goes into it or the challenges we face. What does that mean? It means ensuring that we set the objectives upfront. We have it written down and approved by the other team. That means setting up regular communication to discuss progress and challenges. One of the hardest things to do is to come to a client when you hit a roadblock (maybe the data is not accessible, or it is more expensive than you initially thought). We have always found it best to have those hard conversations upfront and be transparent about the solution we have to offer, the challenges we encounter, and ways to bridge that gap.
On Data Science For The Government Sector
The biggest challenge for the government sector is finding the right talent. It is difficult for them to be competitive in terms of salaries, although there are many great benefits in job security and growth. Recruiting remains a big challenge in getting data scientists into the federal government.
Another challenge is not necessarily the lack of investment because, at this point, everybody (especially across the federal government) understands how important data is. It is not a lack of wanting to learn. It is just that there can be barriers to finding the right resources and getting the time you need to actually learn these skills. People in the federal government are very busy. They have a lot of work to do, and it can be tough to get away.
I am heartened to see the number of policies that have come out recently regarding the importance of data literacy across the federal government. Seeing that becomes such a priority will allow people to become more comfortable taking the time they need to get those training skills. We trained many folks in the government, and everybody we have trained is excited to take the skills we have taught them and turn that into helping their constituents. For us, it is a win-win to support those efforts.
On Building meldR
I am really excited that we have launched our first product ever that specifically addresses the need of healthcare organizations. It is very overwhelming for them with data as there are a lot of different systems where the data exists. For us, we saw a big gap where healthcare organizations need to upscale their employees. When we delivered these types of programs, we started to get many requests about how to build a community beyond the classroom: How to foster mentorship programs? How do your teams know what training programs are coming up? How to market that to get them to join?
So we saw all of these pain points about community building, skills tracking, and communication. We said: “Just like we did seven years ago, if we cannot find a solution to this ourselves, let’s build it.” That is where the idea for meldR came out. We saw an opportunity to bring in our learning expertise as well as our technical expertise to develop a platform that can help (1) foster communities via mentorship programs and communication platforms and (2) schedule additional events outside of the classroom to make sure that the skills are being used.
On Bootstrapping Data Society
Source: https://www.kickstarter.com/projects/978832055/data-science-for-everyone
The main challenge is ensuring we have more money coming in the door than going out. It was easier when we were just two people with a laptop since we did not have many expenses. But then, as we grew, we needed to ensure we had the revenue to match it. From that perspective, we were pretty conservative about hiring at the beginning. We wanted to ensure we could only bring people on that we would be able to keep with us. In fact, our employee number one has been with us for five years. We see that as a testament to how we have managed our finances because we do not ever want to let somebody go due to a financial crunch.
Beyond that, it was just ensuring we had the work for it. That is why we were so visual about the available opportunities so we could always have money coming through the door. That is just a lot of hustle with many conversations and connections.
On Shifting From a B2C to a B2B model
In the early years, we were focused on B2C. We were doing fine, but we noticed that in order to be truly competitive in that space, we had to spend probably millions of dollars on marketing. As a bootstrapped company, we just did not have that kind of money lying around. That was one of the factors in our decision. Then the other factor: as we were delivering these virtual instructor-led courses, we would get requests to train teams. This is an area where we can build relationships and further our mission of empowering people to use data. We thought, in order to succeed, sometimes you have to do less better. We obviously had limited resources, so we abandoned the B2C revenue and went after the B2B revenue. At the end of the day, that was the better option for us as a company.
Some of our early clients were the Department of Commerce and Northrop Gunman. Most of the early deals we won via referral. Success begets success. Once we did a good job in the earlier programs, we got more referrals. It is important not only to say that you deliver a good product but actually deliver a good product.
On Hiring
Hiring is absolutely the most critical aspect for any company, especially for a small company where every person has an outside impact. For us, we found success with people who do not necessarily have a traditional background (like an undergraduate degree in mathematics) but who have taught themselves those skills out of their passion. They love solving challenges and take responsibility for the projects they are working on. As a small company, there are pluses and minuses, but the biggest advantage of joining a small company is that you have the ability to create change on a larger level. We always ask for the opinions of our team members because we do not have a monopoly on the best ideas by far. I always like to hire people who are smarter than me. I should never be the smartest person in the room. So again, we look for people who are excited about teaching, love learning, work independently, and treat others with respect.
On Bringing More Goodness to the World
As a small business, we are now at a point where we are able to give back to our community monetarily. In early 2021, we partnered with Rise Against Hunger, an organization to help food insecurity around the world. We also partnered with World Central Kitchen to help them access food in hard-hit areas. There are so many people in the US who do not know where their next meal is coming from. That is an issue everyone can rally behind. We are looking at other ways to continue with these initiatives, maybe doing another partnership with Rise Against Hunger or volunteering with organizations like Martha’s Table.
On Engaging More Women in Data Science
The best way to engage more women in data science is to be a role model and be available. DC is one of the best places to be a woman in technology. There are so many women who have been incredibly helpful to me in my journey. They have given me their time, relationships, resources, etc. That level of encouragement has helped me get to where I am today. For me, being a role model for anybody who is thinking about getting into the field is very powerful.
The other piece is about having events where you can help people understand the basics of data science and programming. Many people without experience in data or programming feel a lot of fear and trepidation about even trying this right. One of the biggest challenges we face in training is overcoming that barrier to entry. Especially as women, we are told as girls that math may not be our subject, so we internalize that. As a result, if you look at the trajectory of women in technology, it is going up a bit now. For a while, women were very strong in math and science until about middle school. Then you see that trend decreases in high school. It is not because of our ability. It is because of what we are being told. Being able to show them what data science is and get them comfortable with it will help propel them into the next learning phase.
On The Evolution of Education
I think over the next 3–5 years, we are going to see a lot more evaluation of equity. How can we ensure that people have the resources they need to succeed? Some initiatives are happening now to get fast Internet across the nation. We will see a rise in K-to-12 virtual learning and maybe homeschooling. Some parents cannot wait to get their kids back in the classroom, and I think maybe others have found a way to make that work — potentially more individualized instruction through small groups.
Hopefully, people will be nicer to teachers, maybe give them some more presence, a bottle of wine, and some gift cards. As somebody in the classroom, I know the past few years have been very difficult. Without high-quality educators in our schools, we are doing a disservice to our children. It is important more than ever that we have the right people in the classroom who can feel passionate about teaching.
Show Notes
(02:18) Merav talked about her undergraduate experience at McGill University studying Psychology and Sociology.
(04:33) Merav discussed important attributes of an exceptional teacher given her two years teaching elementary special education in NYC public schools through the Teach For America program.
(08:19) Merav commented on her time working at the International Baccalaureate Organization and working as a Kaplan GRE instructor.
(10:57) Merav shared the backstory behind the founding of Data Society, a predictive analytics training and consulting company (co-founded with Dmitri Adler and John Nader).
(14:15) Merav reflected on her journey into programming.
(17:16) Merav explained why data science training should be industry-tailored for maximum success.
(20:57) Merav talked about how Data Society creates and evaluates its training curriculum.
(23:59) Merav provided an example of how Data Society provides customized AI solutions to inform decisions, automate time-consuming manual processes, and solve complex data challenges for its clients.
(27:38) Merav brought up challenges that hinder the adoption of data science in the government sector.
(29:49) Merav unpacked the six different steps for organizations to start moving up the data analytics maturity model.
(33:07) Merav dissected meldR, Data Society’s internal product built for Learning and Development teams in healthcare.
(36:24) Merav reflected on bootstrapping Data Society in the early days (look at this 2016 Kickstarter campaign).
(39:48) Merav discussed the shift from a B2C to a B2B model for Data Society and scoring partnerships with Fortune 500 companies and federal agencies.
(42:47) Merav shared valuable hiring lessons to attract the right people who are excited about the mission of Data Society.
(45:22) Merav shared her experience shaping the remote work culture.
(49:05) Merav touched on initiatives at Data Society to bring more goodness to the world.
(50:28) Merav provided different ways to engage more women in data science (via the Women Data Scientists DC Meetup and DCFemTech).
(53:17) Merav predicted the evolution of education in the next 3 to 5 years.
(55:29) Closing segment.
Merav’s Contact Info
Data Society’s Resources
Mentioned Content
Articles
“Is Your Enterprise Data-Driven?” (May 2021)
“Why Data Science Training Should Be Industry-Tailored for Maximum Success” (August 2021)
People
DJ Patil (The first Chief Data Scientist of the US)
Hilary Mason (Co-Founder of Hidden Door)
Avriel Epps-Darling (Ph.D. candidate, Ford fellow, and Presidential Scholar at Harvard University)
Book
Weapons of Math Destruction (by Cathy O’Neil)
Notes
My conversation with Merav was recorded back in December 2021. Since then, many things have happened at Data Society. I’d recommend:
Reading Merav’s articles on Forbes about creating a culture of data sharing, assessing data literacy, and communication in the learning process.
Reading Data Society’s white papers about data science in research and data science in healthcare.
Checking out the Camelsback product for risk assessment in financial services.
Trying out the Data DNA assessment tool for organizations’ data maturity.
Finally, Merav was also just recognized as one of the DC region’s 40 Under 40. The awards are given annually to recognize the outstanding achievements of young leaders in the Washington, DC, area who lead the community forward through hard work, philanthropy, and community engagement.
About the show
Datacast features long-form, in-depth conversations with practitioners and researchers in the data community to walk through their professional journeys and unpack the lessons learned along the way. I invite guests coming from a wide range of career paths — from scientists and analysts to founders and investors — to analyze the case for using data in the real world and extract their mental models (“the WHY and the HOW”) behind their pursuits. Hopefully, these conversations can serve as valuable tools for early-stage data professionals as they navigate their own careers in the exciting data universe.
Datacast is produced and edited by James Le. Get in touch with feedback or guest suggestions by emailing khanhle.1013@gmail.com.
Subscribe by searching for Datacast wherever you get podcasts or click one of the links below:
If you’re new, see the podcast homepage for the most recent episodes to listen to, or browse the full guest list.