It’s officially been three weeks since I’ve started my new role as the data engineering manager at Vox Media.
Other than the work I’ve been doing to feel through ongoing projects and processes, I’ve been having fun articulating and ideating ways to make major wins for our data team. In this post, I’d love to share my thoughts on the importance of Data Science teams and how you can also begin to define a data strategy moving forward.
Why does your team exist?
Our Data team at Vox Media is a small but growing and knowledgeable group of data enthusiasts.
We exist to…
Provide Vox Media stakeholders with democratized and integrated data solutions to gain clarity, insights, and scale on first-party data.
Our vision is to…
Become a world-class data engineering team leading the way in developing production-ready and state-of-the-art data science solutions.
We do this by…
We stay curious, share what we know, support our team members, celebrate our successes, and learn from setbacks.
My first week at work involved meeting all the amazing people making these goals possible. In the world of software technologies, the software solutions we build are meant to be useful, loved, and impactful. Our code doesn’t exist to live in silos, never to be seen or understood. Bringing a team together means understanding that there are so many pieces to the journey of delivering valuable software solutions. And that this journey is cross-functional.
Get Clear on the Details
My second week at Vox Media was defining these differences and articulating them by functions.
Understanding the skills and functions needed to build fully democratized and integrated projects isn’t easy, especially when it comes to building teams (such as data science teams) where there is no industry standard for specific roles. How do we define a Data Engineer vs. a Machine Learning Engineer vs. a Machine Learning Researcher?
I began articulating these concepts as a means to have more discussions on perspectives that traditional teams across organizations may not realize as it relates to building team compositions that are quite different from the standard Software Engineering team (those as we’ve seen with DevOps and SRE Team Toplogies that even this is ever-changing). My ideas needed to acknowledge the technical contributors of the data team and the subject matter experts that help make our projects and ideas tangible. How do we have all stakeholders begin to understand and respect the priorities, skills, and differences that come with having different roles, backgrounds, and values? And What does this look like when the sum of the parts comes together?
I began to define these pieces and clarify as time went on. There are examples of amazing data science teams that also give back to the community through research publications, available datasets, and open-sourced solutions. Facebook Ai, LinkedIn, and Mozilla all have great data science teams. If our vision is to become world-class, there are certain and standards to be met.
I come from a DevOps background, where we’re often stuck in (sometimes transient) terminology, asking each other what is DevOps, What is SRE, What is GitOps, or What is Progressive Delivery? But, if we can focus on the buckets and areas of expertise, we can begin to have the right roles and capabilities for different projects on our team. I strove to also think about the current talent gap in the data science space, how we have over 21 million software developers in the world, but not nearly as many that understand and work on data solutions.
I ask these pointed questions not only in this post but also to my community and team. These discussions and conversations are what make new teams more well-rounded, inclusive, and successful.
I’ve come into the organization at a time when we’re so curious and excited about data and ethical usage of data in modern platforms and solutions, such as Forte. However, this doesn’t mean anything if we can’t fully articulate the value of our plans at an inter and intra-team level. Communication runs into a scale problem quickly when team dynamics are set in a top-down hierarchy.
I shared our team’s vision and mission statements earlier in this article because I am passionate about these outcomes. Sharing and defining this with my team was one of the first actions I took as the new manager. In modern software development, autonomy is how we can sustainably and reasonably scale to meet software demands. The contributors decide what to work on and when to get it done. As a manager, I’m here to support as best as I can.
I hope this was a great introduction to some of my reflections on what it takes to build a data science team. Coming into this role was by no accident. I’m passionate about bringing teams together to deliver their software with confidence and success. I do this by sharing what I know, holding space, and inspiring anyone interested in tech to do the same. I’m Tiffany Jachja, a Data Engineering Manager at Vox Media, and I’m excited to bring Data Science solutions to our networks, viewers, communities, and industry.