It was great to meet Jerrie in person at dbt Labs Coalesce 2023, and furthermore, to find out it was her first day as Resident Architect at dbt Labs 🙌.
Jerrie has given some great talks, including this one at MDSFest, and has a wealth of data experience across both public and private sectors, which makes her a unique guest in this series.
Intro
I am a creative problem solver who has always been driven by a desire to understand what motivates people and how I can influence the world around them.
When I first ventured into tech, it felt like starting from scratch. After dedicating eight years to the non-profit and public sectors, I transitioned from the realms of social sciences and public health to what was portrayed as the dreamlike world of tech and big data. At the time, it seemed as if I was stepping into a sci-fi universe, ready to harness newfound powers. However, in reality, many challenges remained consistent, but the lens had shifted. Instead of focusing on macro-level issues—like shaping behaviors to enhance public health and reduce mortality—I was now honing in on micro-level concerns, such as refining individual products, systems, and initiatives.
[David Jayatillake - I would like to highlight work outside the private sector as worthwhile to data folks - most people who read this blog don't work in the public sector. Can you share some pairs of public/private sector work, where the method is more or less the same, but applied differently for the two?
Jerrie Kumalah - I worked on a lot of collective impact (https://collectiveimpactforum.org/what-is-collective-impact/) initiatives. A major part of these initiatives is creating a shared vision, clear measurements that are the equivalent of company KPIs that everyone agrees on, priorities, and a plan of action. The process is very similar to what it takes to put a product roadmap together. You have to create a vision, define priorities, have short and long-term goals, and have a clear plan of action. For products, the scope tends to be more defined and very specific, whereas, in the public sector, the issues are not always as well-defined - that can become a challenge when developing action plans. A lot of the approaches we use tend to be informed by our industries. The reality is that there are only so many ways you can do things, so a lot of these concepts are a matter of translation and looking for what theories of practice are relying on.]
My education and background
I studied health and societies and African studies during my undergraduate years. I was keen on understanding human behavior and history, and how these influenced cultures and health. I explored many majors before landing there. I was that student who explored many majors. I started in the fine arts and wanted to study animation, but, being a child of immigrants, it was frowned upon. I ended up giving architecture a try, which felt like a good compromise between art and the structure my parents wanted. However, that didn’t seem like a good fit, so I decided to focus on something that would allow me to work with people and help improve the state of the world. I went on to get a graduate degree in health behavior and from there, started a career in public health.
When reflecting on my data career, I used to begin with my first job with a data analyst title and not my first job post-public health degree. However, looking back over 14 years since graduate school, I realize that I've always been immersed in the world of data.
I started my career gathering primary data: conducting interviews, collecting biological samples, and observing and coding behaviors for a comprehensive study on the influence of environmental factors on birth outcomes. What I learned very early on is that how you collect data significantly impacts the story you can tell. I was immediately fascinated by the power of data to shape the world around us.
I went on to work in several roles in the public and non-profit sectors. I’ve helped organizations build data sets, use their data in creative ways, and dig into publicly-available data sets. The focus at the time was to shape programs, invest in communities, and inform policy decisions. The common thread across these roles was direct engagement with the people who needed this information to make decisions and take action. I was on the ground talking to people, teaching others how to use data to better tell their stories, and being an advocate for using data as a tool.
While these roles were rewarding, I noticed a rigidity in approach and a lag in adapting to evolving data trends. Eager for flexibility and innovation, IÂ taught myself how to code in R and leaned heavily on my public health biostatistics background and my ability to understand and decode complex problems.
I then went on to work at a marketing firm, building forecasting models. I was one of their very first analysts, helping them understand how they could put their historical data and myriad of spreadsheets into a codified source of knowledge that truly captured their expertise and the lessons learned over time. I then worked on automating reporting for the fundraising team at Wikipedia. Then, as a data scientist building graph models for a startup, redefining how we could connect disparate datasets in the public sector to better show interactions across social services. To a full stack analyst turned analytics engineer, using all my skills to create data products that would shape my company's products, systems, and operations.
[David Jayatillake - Did you feel less fulfilled working in the private sector? It's easy to end up driving the wrong things with data in the private sector... "buy more stuff you don't need", "borrow more", "gamble more". Did you feel like this wasn't present in the public sector type of data work?
Jerrie Kumalah - In the public sector, it is easy to get complacent and believe that whatever you are doing, no matter how small, matters - but that comes at the risk of not going outside your comfort zone and holding yourself accountable when things don't generate the value or results they are supposed to be generating. I loved the passion that people had, but I often struggled with the lack of urgency and limited speed to innovation.
In the private sector, there is speed but there isn't always the right level of intentionality. I've noticed that strategy - meaning clear long-term planning, strong collaboration, and focus - is not always a strong point. Every team can easily become an island and the big picture can easily be lost.
What I've always been strong at is working with people and making sure I create a shared vision and focus on whatever work it is I am doing. That has been the main driver for me across both the public sector and private sector and has become the main thing I focus on. I always try to think in terms of - How can I bring focus here? How can I make sure we are on the same page? How do we hold ourselves accountable?]
I’ve seen so many different spaces and prototyped my way to where I am now. Learning, growing, problem-solving, and pushing myself outside of my comfort zone. I’ve seen everything from homegrown data tools to data stacks that many of us would be envious of. Despite the variety, a common challenge persists across all these sectors: the human element.Â
The questions remain:
How do we show our value?
How do we collaborate effectively?
How do we focus on the right things?
Human interfaces - the good:
Earlier on in my career, I worked as a CitiStat analyst. My job was to work with my assigned city departments and help compile the data they had available to evaluate performance and identify opportunities for improvement. The data was reviewed biweekly and informed how resources were invested in the department.Â
What I learned:
Accountability matters. People need to feel some kind of pressure to stay focused on the right things
A clear process is important. Reviewing metrics regularly can help shift priorities and investments to better align with your vision for success
This was not a job where I sat at my desk all day. A big part of my job was being in the field, learning about the workflow, investigating how data was generated, determining how data collection could be improved, working very closely with department leaders to get their buy-in, and making sure that all my work also helped them make better decisions. I was both a diplomat and an enforcer. I visited rec centers, sat in on trash truck routes, followed city arborists counting trees, and even investigated trash dumping across the city and how we captured that.Â
[David Jayatillake - These personas - diplomat and enforcer - have skills that not all data folks have. Did you have them at the start of you career? Did you learn them? How did you acquire them?
Jerrie Kumalah - That is a fantastic question. I've reflected on this one a bit over the years and a big part (larger than I cared to admit before) has to do with my upbringing. I grew up across 7 different countries, 3 languages, and 10 different schools by the time I hit high school, so I learned early in life the art of bringing people together and truly listening to understand and adapt. This is a skill that is second nature to me and I have used it everywhere that I have ever worked. I listen with the intent of understanding how people view a problem and how it impacts them.
I am a firm believer that accountability is the true lever to generate any value. If you are working towards a goal, you need to define the risk for not meeting it or the reward/impact for achieving this goal. But the hardest part is having someone who is not afraid to remind people what you are working towards. We like to assume that everyone is on the same page, but this is NEVER the case. When I started noticing that there wasn't always someone taking on that role, I started asking the questions that would help bring focus whenever possible. It has taken practice and, by taking on that role, you quickly learn which teams are really ready for these conversations and which ones are not quite ready for true accountability.]
We all like to believe that people come with the same focus, but the reality is that everyone has their own agenda. This is why the people involved in a project matter and why you need to take the time to develop the systems and processes that will lead to sustainable impact.Â
What is sustainable impact?Â
Accountability
Planning
Collaboration
Clear objectives
My favorite part about my role as a CitiStat Analyst was that the departments were ultimately accountable to the mayor and the people of the city, so, whether they liked it or not, they had to stick to the process. But, for it to work, we had to focus on what was most important to these departments and take the time to truly understand the work, the pain points, and how we could use data to better tell their stories.Â
Human interfaces - the bad:
Scenarios that I have often seen:
There are no owners, so there is no real collaboration and there is nobody accountable for the result. In this scenario, nobody wants to take it on because there are no incentives to do so or it falls into a category that is often a need, but not part of any one business function.Â
Everyone is an owner but nobody is responsible for actually making the final decision, so nothing gets done. In this scenario, collaboration is often strong but accountability is weak - nobody wants to be the one pushing and asking the hard questions.
There is a decision maker that makes decisions with absolutely no input from the people it impacts the most. It feels disheartening and limits collaboration and willingness to invest in solving the problems in front of you.Â
What I have seen work best is when collaboration, accountability, and clear objectives are defined and implemented.
When I worked at Firefly Health, we would meet quarterly as pods, which included product, engineering, program, and analytics staff and we would have to agree on the top priority that we were all focusing on. That priority would be the number one for every team and everything else would be secondary or in support of maintaining things that were needed to maintain business as usual. What I loved about that process was that every team had to make sure they defined how their work would tie back to that priority.
As the data rep, I made sure that everyone considered how they could use data to inform, shape, or determine if they were going in the right direction. The approach allowed data to be embedded across each function and not just as a siloed element or afterthought. I worked closely with engineers to ensure that the features they were building met the needs of the product teams, by proactively defining and determining what data points were needed. Getting to an agreement was sometimes time-consuming, but the process fostered collaboration and added clarity for the quarter.
[David Jayatillake - How did you get to have this level of participation in the product and planning process? Was it something that was already present in the company, or did you have to use some level of force/persuasion to get in?
Jerrie Kumalah - I was lucky enough that the planning process was already part of their structure, which really made it easier to collaborate. What I had to work on was helping shift how they thought about data: moving away from using data reactively and starting to think of it proactively and as an essential part of shaping our products and operations - not just as a reporting mechanism.
This meant talking to each team lead, prototyping ways in which it improved how decisions were made, and including myself in meetings where data had not always been included in the past.]
What are you doing now, in your current role as an analytics engineer, that helps make human interfaces in data better?
These past 2 years, I have been working as an analytics engineer and one of the things I spent a lot of time focusing on is creating a shared understanding. In practice, this means clear documentation, project plans, and getting to know my stakeholders. I have not relied on fancy tools or processes, I just tend to focus on the people and making sure that what I’m working on is generating the value they need now. You can find me talking about these concepts and you can hear some of my thoughts here, here, and here.
I spend a lot of time thinking about how we get people excited about data and how we stay focused on solving problems, instead of just getting enamored by every new approach and new tool. I just joined dbt Labs as a Resident Architect on the professional services team, so I’m on to a new adventure in how to use analytics engineering concepts on larger data problems.Â
[David Jayatillake - Why have you chosen to become an analytics engineer? Do you feel this is a material change from your previous roles, many of which would be considered analyst roles, if I'm correct?
Jerrie Kumalah - I fell into analytics engineering in my quest to hone my data problem-solving skills. I saw analytics engineering as the perfect cross between building the data systems and the infrastructure needed to make data usable and still being closely connected to the business and the people. Although I've loved analytics engineering, I've also noticed a trend where we are moving further away from the people/business and getting so in the weeds that the focus is on the tools and systems and not on the problems we are solving.
I recently accepted a role as a Resident Architect at dbt labs. So now, I'll be shifting my focus to thinking about how I can help a company shape its data systems, based on the tool they are investing in. It will be a combination of finding the right technical solutions, understanding stakeholder needs, sharing best practices and, I'm sure, a myriad of other things. I'm excited to see what this next chapter teaches me. For me, it is all about solving problems and learning from the many different data roles we have in the profession, so I'll keep prototyping my way to being the best data problem solver I can be.]