Data-Driven Work Cultures: Sarah Nagy Of Seek AI On How To Effectively Leverage Data To Take Your Company To The Next Level
An Interview With Fotis Georgiadis
Data engineering is the foundation that must be laid in order to start getting ROI out of your data, and making the first data engineering hire will set the tone for your data infrastructure for the next several years. I have come across a few companies that made data engineering hires that chose lower-quality vendors or built the data engineering infrastructure with an inefficient architecture. When these mistakes are made at the foundational level, they compound as the rest of the data stack and data team grow. Sometimes, new data hires are needed just to put bandaids on the inefficiencies of this foundation, and it can get very expensive to start all over at this point.
As part of our series about “How To Effectively Leverage Data To Take Your Company To The Next Level”, I had the pleasure of interviewing Sarah Nagy.
A former data scientist, Sarah Nagy founded an analytics automation startup, Seek AI, in September 2021. Sarah most recently led the consumer data team at Citadel’s Ashler Capital, and prior to Citadel, led the quant arms at two successfully exited startups and developed algorithmic trading strategies at ITG. Sarah has a Master in Finance degree from Princeton and dual bachelor’s degrees in Astrophysics and Business Economics from UCLA.
Thank you so much for joining us in this interview series. Before we dive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started?
I started out as a researcher in astrophysics at UCLA and Caltech, working with data from the Hubble Space Telescope. When I learned about quantitative finance, it seemed right up my alley so I pursued my MFin degree from Princeton and went on to work on Wall Street doing algorithmic trading. I had always been interested in startups, though, and between that and the growing field of machine learning, I decided to work as a quant/data scientist at a couple of startups. I’m fortunate enough to have worked alongside the founders of those startups so that I could learn from them and apply these learnings to Seek.
Can you share a story about the funniest mistake you made when you were first starting? Can you tell us what lessons or ‘take aways’ you learned from that?
The first day I changed my LinkedIn status to the CEO of Seek, I unsurprisingly received a lot of messages from friends and colleagues, but also some messages from people I did not know. One message that made it into my orbit was from a production company claiming that an A-list celebrity (one that I was a fan of) wanted to interview me about Seek. After doing a bit of research, I learned that the “catch” was that I’d need to pay a five-figure sum to get to interview with the celebrity. Not only that, but I had wasted a couple hours of my time doing research on an opportunity I didn’t even need at that early of a stage, when I needed to focus on… building the product. This experience was the beginning of me starting to block out all the noise surrounding me as a startup founder and focus on the things that really matter.
Is there a particular book, podcast, or film that made a significant impact on you? Can you share a story or explain why it resonated with you so much?
I was really inspired by Dorothy Vaughan in Hidden Figures, teaching herself to program the IBM computer and, in doing so, arming herself with this new technology. As a data scientist writing manual code a significant amount of the time–often to answer data questions for less technical colleagues–I frequently felt like a “human computer” working with outdated technology. My prediction is that AI will arm data scientists/analysts with the ability to avoid this manual work and instead be able to focus on the projects that require deep focus and can add a lot of value to the business.
Are you working on any new, exciting projects now? How do you think that might help people?
We have some very exciting new integrations that we will be releasing early next year in the Seek platform, which I am really excited about. Giving our customers more ways to embed Seek into their workflow is really exciting to me, knowing that I am making it easier for our end-users to access the data they need.
Thank you for all that. Let’s now turn to the main focus of our discussion about empowering organizations to be more “data-driven.” For the benefit of our readers, can you help explain what exactly it means to be data-driven? On a practical level, what does it look like to use data to make decisions?
To illustrate what it means to be data-driven, I’d like to point out an observation from my career working with data: some organizations may work with data, but not be data-driven. I remember the first time I encountered a research firm that took a traditional approach to working with data, as opposed to a data-driven approach. I was shocked to hear the analyst tell me about a stock price prediction that was first made based on qualitative, not quantitative, reasoning. From there, the analyst had manipulated the data to fit this qualitative prediction. I was perplexed to see this, as it seemed to invalidate the point of working with the data in the first place. Yet, I think that many organizations may encounter this pitfall without embedding their data team’s expertise in business users’ interactions with data. How to make this scalable is certainly a challenge, but the scalability is increasing as data tools increase in sophistication.
Which companies can most benefit from tools that empower data collaboration?
In most business-facing roles, performance is becoming more and more correlated with data accessibility. The higher this correlation is in an organization, the more data collaboration is needed. Companies with the highest correlation include certain verticals–for example, many B2B SaaS businesses live and die by their data. Say a salesperson needs to ask 20 ad-hoc data questions per week in order to accomplish their goals. The better and faster the collaboration with the data team, the more of these questions can be answered, and the closer the salesperson can get to their goals.
We’d love to hear about your experiences using data to drive decisions. In your experience, how has data analytics and data collaboration helped improve operations, processes, and customer experiences? We’d love to hear some stories if possible.
In the algorithmic trading world, from my experience, there are few ways to avoid the truth and win. This is what drew me to algorithmic trading coming from academia, where I was used to letting the data tell me what was true, and not the other way around. As hedge fund manager Ray Dalio often says, “Getting to the truth is essential in order to win.”I believe this applies to every data-driven business, not just in the financial sector.
Working at certain companies, I was appalled to see business intelligence tools show bad data to the business-facing users, who knew little about the underlying data. These business users could have made detrimental decisions for the business by trusting this bad data. Something I think that businesses will realize in the coming years is that giving accessibility to their data is not enough; they must make sure that the data is right.
Has the shift towards becoming more data-driven been challenging for some teams or organizations from your vantage point? What are the challenges? How can organizations solve these challenges?
Similarly to my last answer, one of the biggest challenges for organizations is managing the tradeoff between accuracy and accessibility. On one hand, accessibility allows less technical folks to start interacting with the knowledge wellspring that is a company’s data. On the other hand, what good is a wellspring of polluted water (i.e. bad data)? This is the other side of the tradeoff, which is making sure the data that business users see is accurate. The best data teams are those that manage this tradeoff in the most optimal way possible, and a big part of that is carefully calibrating and vetting any tools that non-technical users can interact with.
Ok. Thank you. Here is the primary question of our discussion. Based on your experience and success, what are “Five Ways a Company Can Effectively Leverage Data to Take It To The Next Level”? Please share a story or an example for each.
In the order of the most basic to the most advanced:
- Move to a cloud data warehouse.
Despite encountering them frequently, I am always amazed when I meet with companies in 2022 that don’t have some sort of cloud data warehouse. Having entered the workforce in 2014, I have spent my entire career working with cloud-based tools and sometimes forget that there are other options. But, when I encounter alternatives, it is painful to see. The biggest pain point is all of the maintenance that goes into maintaining these databases, when cloud data warehouses do all of that maintenance for you at a low cost so that you can put that money towards something with a much higher ROI.
2. Hire a good data engineering lead.
Data engineering is the foundation that must be laid in order to start getting ROI out of your data, and making the first data engineering hire will set the tone for your data infrastructure for the next several years. I have come across a few companies that made data engineering hires that chose lower-quality vendors or built the data engineering infrastructure with an inefficient architecture. When these mistakes are made at the foundational level, they compound as the rest of the data stack and data team grow. Sometimes, new data hires are needed just to put bandaids on the inefficiencies of this foundation, and it can get very expensive to start all over at this point.
3. Get analytics engineers.
I am really excited about the future of the transformation layer, which is a system that abstracts difficult-to-work-with raw code into simpler, more business-focused code and, in doing so, makes it much easier and faster to get insights. The purpose of an investment in a data team and data stack is to generate an ROI, and this ROI comes from insights that are extracted from the data using code like SQL. This means that the transformation layer will allow organizations to finally get tap into this ROI that they have spent so much capital building towards. We work with many companies that are early adopters of the transformation layer, and the productivity it has given them enables them to explore other high-value projects that their competitors wouldn’t have time to do.
4. Make your data accessible to non-technical users.
Continuing from the above, what good is data if it’s not generating an ROI–and how can a business get this ROI if the business-facing users can’t even see it? This is why it is absolutely essential to give access to as many people as possible (without compromising accuracy, which I will get to in #5). This is where the ever-so-buzzy term “data mesh” can be helpful. It may be buzzy, but I agree with its premise of distributing the data team across an organization to be able to support more end users. While these distributed data teams are smaller than one central data team, with the right automation tools, this type of structure can be incredibly efficient at giving more end-users access to the data they need.
5.Scale your data team’s knowledge.
Regardless of the structure of your data team, you will encounter the challenge of scaling the data team’s ability to serve the end users. Some of this scalability can be achieved with thoughtful conversation around which manual work to prioritize. For example, determining which dashboards or semantic models take the highest priority to build is an essential conversation to have. Another option is to consider automation tools coming on the market, which can store and replicate the data team’s ability to produce these deliverables on-demand. This type of automation can elevate the data team from the depths of manual work to the overseers of a more scalable and sophisticated architecture.
The name of this series is “Data-Driven Work Cultures”. Changing a culture is hard. What would you suggest is needed to change a work culture to become more Data Driven?
To change any work culture, no matter what it is, it always starts from the very top. If you’re a CEO trying to change your company culture to be more data-driven, the first thing I’d recommend is to make sure you yourself are data-driven.
Being data-driven doesn’t necessarily mean knowing how to code, but it does mean knowing how to ask the right questions and interpret the resulting charts and tables. Also, asking follow-up questions can be a very powerful way to hone in on the incredibly niche insights that can really change your business the most. As a data-driven CEO with absolutely no time to code, I simply ask the questions that I have without worrying about the code, and once I get the answers, I make sure to ask follow-up questions to confirm what I think I’m seeing or learn more about any insights that stand out to me.
Once you are comfortable with your own skills working with data as an end user, try showing your colleagues what worked for you and what didn’t. As the saying goes, “When you change yourself, you change the team.”
The future of work has recently become very fluid. Based on your experience, how do you think the needs for data will evolve and change over the next five years?
As more data gets collected from more and more sources, its complexity increases. While this is happening, a data-driven culture will go from a competitive advantage to table stakes, and the real competitive advantage will be in getting the most valuable insights the quickest. I predict that at some point, automation will be needed to accommodate all of this complex data and also provide companies with the efficiency they need analyzing it all.
Does your organization have any exciting goals for the near future? What challenges will you need to tackle to reach them? How do you think data analytics can best help you to achieve these goals?
As a startup, we are growing fast and constantly improving the product. Collecting the right data in order to help the right customers and build the right features into the product is absolutely essential, and the only way for us to hit our ambitious goals is to have a data-driven culture. Our data pipeline is ingesting data from an increasing number of sources that we can use to quantify how well we’re hitting our goals and also shape the future strategy of Seek.
How can our readers further follow your work?
The easiest way is to follow me on LinkedIn or email me at [email protected]. I love meeting new people and nerding out about data. Please drop me a line!
Thank you so much for sharing these important insights. We wish you continued success and good health!
Data-Driven Work Cultures: Sarah Nagy Of Seek AI On How To Effectively Leverage Data To Take Your… was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.
