This is the story of Priscilla Cole, and what she did when she discovered that her ambitions were bigger than the tools she was using!
Connect with Priscilla here! https://www.linkedin.com/in/priscilla-cole-5892549/
Recommended Listening
The Way You Talk About Your Skills Is Costing You Money
Geospatial Consulting As A Business And Career
Getting Where You Want To Go In Your Career
Applying For A Job, Getting Picked and Negotiating
Mentorship Leadership And Career Advice
In Conversation
Falling Into GIS
Daniel: You told me you kind of fell into GIS — you didn’t know it existed. Can you walk me through that?
Priscilla: I started off as a biology major focused on ecology, which I now think of as inherently spatial in nature — habitats, species distributions, natural resources, the interplay of all these things — but geospatial wasn’t part of that education. Later, in an interdisciplinary master’s combining policy and water resources, I met another student who had maps on his computer screen, zooming in and out, looking at corn in the Midwest from USDA data. I’d never seen data up on a screen like that. He said I should really take a GIS course, and remote sensing too. So I took one GIS class — an Esri-based class, so ArcMap and all of that — and one remote sensing class. That was it, two classes, all because a friend said “you should really do this.”
Daniel: After those two classes, were you hooked — or was it too difficult?
Priscilla: The geospatial software in the computer lab was really clunky — it collapsed all the time, the network went down, computers blue-screened, and the data management was difficult with no background in it. I really enjoyed the projects, but it was so much work. Interestingly, as much as I liked it, I didn’t incorporate any geospatial into my thesis on water resources at all — it just didn’t occur to me how I could. I was really into it, but I still wasn’t sure what to do with it.
Drinking the Esri Kool-Aid
Daniel: How did you get from not using GIS in your thesis at all to getting Esri-certified?
Priscilla: In grad school I picked up an internship at an estuary program working on water resources, which turned into a fellowship and then a full-time job. There was one Esri license there, funded by a grant somebody had written, and when I came on staff they said if I wanted to use it I’d have to take over the technology grants and renew it every year. Reading that grant more closely, I realised I could get classes for free, books for free, and get myself to the Esri International User Conference in San Diego every year — all for free. So I milked it for everything it was worth: I maxed out the classes, got all the books, attended conferences for years. I drank the Esri Kool-Aid really hard, and took that path all the way to certification — a certified desktop analyst, and eventually the GISP, the GIS Professional certification. I did the math on the GISP: you need either five years as a rock star in GIS or about ten years of playing at it. It motivated me to keep building my resume, presenting at conferences, picking it all up.
Hitting the Wall: Why Learn SQL and Python
Daniel: The world was your oyster doing analysis. Why move away from that into engineering?
Priscilla: At some point I hit a really hard wall. With all these different data sources — vector, raster, LiDAR, data coming in from field staff — our data was all over the place, on servers, with no semblance of a database, scattered and messy. Esri products let you do some data manipulation, but it’s not great, and I realised I didn’t have the basic tools to wrangle the data into one place and reshape it. Big datasets from satellites or LiDAR get so big so fast. I’d really tapped out what I could process with the out-of-the-box Esri tools. That led me to want to learn SQL — the language of the database world — and database design and architecture, and Python, to extend the toolset, because I often had ambitions a lot bigger than the tools in front of me.
Daniel: Why SQL and Python specifically?
Priscilla: At one of the many Esri conferences they made a grand announcement that Python was easier than ever in Esri products, with a new library called ArcPy, and they pushed it really hard. And a light version of SQL was already baked into Esri. So initially my thinking was just that these would help me use the software better and extend its tooling — they seemed like the natural languages to learn.
Teaching Herself to Code
Daniel: How did you learn — courses, or self-taught?
Priscilla: I took an ArcPy class through one of those grant-funded classes, but mostly I just reached out wherever I could. I bought books with titles like “Teach Yourself SQL in 20 Easy Steps,” the thick O’Reilly Python book, and worked through all the tutorials and exercises. It was really scrapped together from whatever I could get my hands on. I bought a Microsoft T-SQL training bundle through Groupon and worked all the way through it. During the pandemic I needed an activity to keep my sanity in lockdown, so I enrolled in a Codecademy data science professional course — an hour or two a day locked in a room, working on SQL, Python, statistics, machine learning, natural language processing, and a lot of data cleaning.
Daniel: Harsh question — do you think anyone actually cares about that certification?
Priscilla: It was something to talk about in interview processes — I could point to the work I did, and I quickly came back and applied those things, building out models with data I had access to, so it was more than just the coursework. It’s also helpful for renewing the GISP, where you have to report all the activities you’ve done and put in screenshots of any certifications.
Learning on the Job (and Crashing the Network)
Daniel: You were applying these things in your day job. Did it always go well?
Priscilla: After the estuary program I moved to New York City — I had a craving to work somewhere with enterprise-level databases, real DBAs, seasoned analysts. I got a job at the New York City Department of Environmental Protection, essentially the water and sewer department, which had hired a whole bunch of analysts and technology people after Hurricane Sandy. I was super nervous, joining what felt like a well-oiled machine, and I studied hard beforehand to keep up. But when I started, a lot of the work was very manual and inefficient — there were processes where everything stopped and all the analysts worked full-time processing things by hand for a week. I looked around and thought, we can clean this up with Python, automate these workflows. My boss said sure, if you can get rid of this awful task. My first iteration kept hitting the network over and over — I crashed our network a couple of times and had the deputy commissioner march over to my desk: “are you the analyst that took down the network?” But I learned the hard way a lot of really practical software engineering practices — how to reduce network traffic, how much to hold in memory versus on disk.
Crafting Your Own Role
Daniel: This sounds a lot like data engineering.
Priscilla: Yeah — there’s a popular title now, analytics engineering, and that’s a lot of what I was doing: automating workflows, building data pipelines, getting data into shape and into the right place so it’s where the analysts need it. I sometimes think of myself as a therapist for other analysts and engineers — the engineers I worked with at DEP were all civil engineers — where I’d sit with them and really feel their pain. Where are the bottlenecks? What do you really want to do but don’t have the tools for? If you start with “let’s solve your problem,” they may not even have their requirements laid out — they can just show you how frustrated they get. Then you go back to the drawing board and figure out what a solution looks like.
Daniel: Did you ask for permission, or forgiveness?
Priscilla: In the beginning at DEP it was more like asking for forgiveness — timidly asking my boss for a bit of extra time, promising I’d get everything else done, tiptoeing around to find space. But after a couple of successes it launched a new phase of my career: I started to be sought out by the civil engineering teams, and even the director of engineering would pull me into pet projects. I developed a niche in that — and it’s not at all what I was hired for. I was hired just to be an analyst, just to make maps. It evolved into a tool-making and data engineering role.
Daniel: What would you say to someone in the situation you were in — they’ve hit the wall and want to do more?
Priscilla: You really don’t have to be bound by your title or what you were hired to do. The world is so open today in terms of training — any time I want to know how to do something I go to YouTube, watch a quick tutorial, and gain a new skill. Code takes a bit to learn, but it’s not impossible, and you can come from a completely different background. You can shape the career you want — you don’t have to wait for anybody to assign it to you or say “okay, you’re good enough now, you can graduate into this.” You can start doing those things right now, today. I hate the idea of people sitting there waiting to be picked, waiting for someone else to recognise them, instead of putting themselves forward and trying things.
Bridging Data Science and Geospatial
Daniel: Would you call yourself a data engineer today?
Priscilla: After leaving the public sector I moved to an insurance company that already had data scientists and data engineers on staff. I was hired for my geospatial skills, and because they felt I could talk to those groups — I understood the languages and the skill sets. They hoped I’d bridge the gap and bring geospatial data into the workflows and processes. I sat somewhere between data science and data engineering. It wasn’t my job to teach people geospatial data — it was to figure out how to incorporate it into their data science workflows and engineering pipelines, enriching the datasets they already used, almost sneakily, using spatial as a join key to pull different datasets together. As far as they knew, it worked like magic.
Daniel: Standing in a room with a bunch of “real” qualified data scientists, I’d have had imposter syndrome kick in. Did you have any of that self-doubt?
Priscilla: I was so excited to take the job — I felt like I’d made it, crossed over into the private sector, was making more money, and got to sit with people who had fantastic skills and were so smart. For me it was like writing the free courses into those grants — I had all these people I could just ask questions of. Fortunately the team was really welcoming, with practices like code pairing, code reviews, skill sharing, and brown bags. Everyone was someone to partner with and learn from, and I loved every minute of building really cool stuff with them.
From Specialist to Strategist
Daniel: What does the future look like for you now?
Priscilla: After the insurance company I went to a data company — a data provider that sold data, including imagery from space showing natural catastrophes, floods, and fires, to insurance companies and governments. I got really sold on the idea that for portfolios with a global reach you need global datasets. I’d learned about remote sensing back in grad school and touched it throughout my career, but I had this aching sense that there’s so much being seen from space, and we need to bring it back down to Earth and make use of it — for humanitarian efforts and natural catastrophe recovery, not just defense, which takes the lion’s share of satellite data. I’ve developed a niche as a very technical person, but I’m coming to understand that one of my greatest powers is having worked across all these verticals — scientists, civil engineers, municipal government, insurance, a data company — which lets me bridge gaps and help people see at a higher level how to bring different fields and data sources together.
Daniel: Do you see that next jump — from technical specialist to strategy — as just as difficult as the move from analyst to engineer?
Priscilla: This is where I do have a master’s in policy, in my back pocket. For years I pushed it to the very bottom of my resume — I never brought it up, because I wanted to be taken seriously as a data person. Now I’m rediscovering it. Many organisations put their data people and engineers into a little box — “you do the data and engineering, we’ll tell you what to do” — with all the strategy decided at a much higher level. I worked so hard to get into that box, and now I’m trying to claw my way back out: I have all these ideas. So I’ve embarked on what I’m calling a sabbatical — I walked away from being employed to do some writing, explore ideas, and work on solo analysis projects. I’ve also taken on organizing a conference, the Geospatial Risk Summit, to bring together people from insurance, finance, supply chain, and the satellite industry. Somewhere in this journey I’ll figure out what I’m meant to do, or at least the next step.
Daniel: Organizing a conference is a brilliant way to position yourself with the right people — you don’t have to know everything about the topic, you just organize the people who do, and you end up at the centre of a group who care about this. Any final advice?
Priscilla: Someone gave me the advice to build your own board of directors, at whatever stage you’re at — you don’t have to wait until you’re starting a company. And build it not just with friends or people who think the same as you — you don’t want to create an echo chamber — but strategically, with people who have different skill sets and different views, combining them into a whole that can really get at things. No one person can do everything; it’s a group effort, and it takes a village to accomplish big things.
