In today’s post, my colleagues Greg, Anna, Anand, Anna, and Robert — who come from a variety of backgrounds ranging from fullstack software development to linguistics, hardcore math, and more — share how they came to work with data at Automattic.
Greg Ichneumon Brown
Greg, you’re an expert in Elasticsearch and you’ve been at Automattic for over nine years. How did you come to work with this technology and at Automattic, originally?
Before Automattic, I was shifting my career after 10 years as a hardware engineer. I had enjoyed designing networking and memory systems, but always had been interested in AI. I quit my job to get a Computer Science Master’s degree with a focus on machine learning and natural language processing.
The project was to deploy Elasticsearch to search the company’s ~100 P2s. (P2s are websites we use to communicate within Automattic.) The project was fascinating. Because I was indexing the primary means the company used to communicate, just by doing the trial, I actually learned a lot about the company culture and history. The free speech and open source ethos of the company was readily apparent.
The project involved replacing a MySQL-based search, so I was replicating WordPress data into Elasticsearch to query it. This was a longer trial project than we do now —maybe 100 hours. I had used Lucene before, but Elasticsearch was only on version 0.16 and I had never used it. Everyone was fairly forgiving of my limited PHP and WordPress knowledge. In the end, the trial went pretty well. The code has been rewritten a few times, but Elasticsearch still powers our internal company search.
Anna Magdalena Kedzierska
Anna, you’re currently working on a natural language processing project. What else have you worked on in your career, and how did that take you to where you are now?
I came to Automattic a couple of years into my transition from science and research into industry. I have postdoctoral experience in Neuroscience, Bioinformatics, and Clinical Trials, and I had been on a path to an academic career in Algebraic Geometry and Statistics. We applied the theory behind Algebraic varieties to problems in Genome Biology and Phylogenetics.
As fascinating as the world of research was, I left academia as the Data Science field was gaining momentum. I traveled the world exploring alternative applications of Math —: from data journalism in South Africa to analyzing NGO data in Beirut and optimizing B2B marketing efforts in London — to name a few. At the same time, I was lucky to be presented with a number of opportunities to hone my programming skills e.g., through the Recurse Center in New York City and the Outreachy grant to work on the Node.js and V8 core. I have always enjoyed coding but this was a game changer and I was hooked (to say the least!).
Anand, you work as a data engineer on our data infrastructure team. How did you become interested in data engineering and what path brought you to Automattic?
I spent the first five years of my career working as a full stack developer. I worked on building interfaces similar to Google Maps in the browser and an isometric game engine in ActionScript. At the same time, I was also working on the backend with technologies like Berkeley DB, Google App Engine, and Cloud Bigtable.
At that point, I was deciding whether to go deeper into one of these areas or become more of a generalist and think about starting my own company. Finally, I chose the first option and joined a company as a data engineer. For the next four years, I worked on building big data products. But because the product team was very removed from product users, I missed the validation of seeing people use the tools I built.
Luckily, I ran across a post about Data Wrangler hiring by Greg on Hacker News. I applied and went through the trial, which in itself was pretty rewarding as it involved learning new stuff and interacting with my future team. Now here at Automattic, I’m working both on our big data infrastructure and on internal analytics tools. It’s pretty satisfying to listen to their problems and solve them. Also, the variety of problems that I get to work on is pretty amazing: I can focus on a specific area to optimize but I can also influence how our whole data infrastructure will evolve over the next few years.
Anna, you lead our data analytics team. When did you start thinking about a career in data analytics and how did you get into this position?
The only child in a family of mathematicians, I originally studied theoretical Economics (does that count as rebellion?), and started a Ph.D in that field as well. I eventually drifted closer to empirical Econometrics; but, as it often happens, I didn’t end up completing my doctorate. A feeling that I wouldn’t be fulfilled by an academic career built in me for a while, and I pivoted to a freelance business — in social media, events, and translation. I missed the clarity, the universal shared language of numbers; but enjoyed connecting the dots between people, my new faster-paced work tempo, and being much closer to the beating heart of business.
After a few years, I found myself working for a global online travel agent in a sales/operations role — and realized I was spending a lot of my time creating reporting and finding ways to access and apply data to drive business and measure success. And I loved it. I eventually became a data analyst full-time. As someone new to data operations in tech, I spent months learning and listening — how are people using data? What are their complaints? What successes do they attribute to analytics, what do they only dream of being able to measure — and how could we work together? I was back to connecting dots in a fast-paced environment, at the root of the business, while seeking (and ever so rarely, finding) clarity in numbers. I’d found my middle ground: a role, an area that I truly thrive in.
I joined Automattic about a year and a half ago, following the recommendation of a friend —now colleague — of mine. I found the transparency, the dedication to open source, and distributed work very appealing, and saw working in a remote team as a fantastic way to grow as an analyst and a professional overall. I now lead the central reporting and infrastructure team I’d originally joined in 2019, and I couldn’t be more excited about our mission to empower business leaders at Automattic to make the right choices about the right things, with confidence.
Rob, you are our newest team member and this is your second stint at Automattic. How did you get into working with data at Automattic, and how did you end up returning to work here?
My path to data was quite winding. As a kid I really liked Math and Physics, but I ended up studying German and then Linguistics. I started learning programming and web development on the side, which I eventually merged into my Linguistics research; I then gradually transitioned from Phonetics to Computational Linguistics, and along the way also picked up a bunch of statistics and machine learning.
I have been a WordPress user and developer since 2006, so I was aware of Automattic for quite some time before joining. After several years working on Natural Language Understanding at a different company, I finally decided to take my friend Greg Brown’s suggestion to apply to Automattic.
During my first stint at Automattic I worked mostly on building out our Hadoop infrastructure and tools to utilize it. I learned a ton, but I missed Linguistics, which is why I left. After another four years working on Natural Language Understanding, I decided to come back to Automattic, primarily for the work culture. I have always believed in the Automattic Creed, and it truly is something that people at Automattic take to heart and practice everyday. This time around I am focusing more on our search products, which is once again in the field of Computational Linguistics, so I get to incorporate both my passion for Linguistics as well as for programming, and I get to help improve products which I care about and use myself.