Rebecca Bilbro, Rotational Labs

Trac Bannon: 

The incredible Jennifer Ives has become a champion for the Real Technologist podcast. She’s been introducing me to an Uber broad set of deep technical experts of all different shapes and sizes. Given Jennifer’s depth in AI and her new company, Watering Hole AI, it came as no surprise when she introduced me to Dr. Rebecca Bilbro. Dr. Bilbro is the founder and CTO of an AI firm called Rotational Labs. She’s also a founder of the popular Open Source library called Yellow Brick that provides visual analysis and diagnostic tools to facilitate machine learning model selection. 

When Rebecca appeared in our Zoom room, I asked her how to pronounce her last name. I was smiling from ear to ear when she said “it’s like the Hobbit, but with an R”. Yep… she said that and my inner geek cosplay toes curled. 

It was 9:00 AM on the east coast and Rebecca had just finished a morning walk with her mom, a retired professor of computer science. I was about to ask her about the walk when she volunteered that they had been discussing data structures. The computer scientist and the data scientist had complementary points of view though Rebecca adds that she thinks of data from the vantage point of being a data consumer. Her origin story was starting to peak out… she grew up in North Carolina as the eldest daughter of two university professors. 

Rebecca Bilbro: 

My parents were both faculty at North Carolina State University, both in the engineering department, just like real technologists from the very beginning. 

Trac Bannon: 

Rebecca is definitely the child of two academics. Her dad was a theoretical physicist who became an electrical engineer. Her mother was one of the first computer scientists having entered the field before it was actually named. Rebecca has her own in-kind experience diving into data science before that term had been coined. 

Rebecca Bilbro: 

My mom was like one of the only women, maybe a dozen women… and she was one of the first computer scientists just because of the chronology it didn’t exist when she was in grad school, so her PhD is in mathematics. But she was recruited into one of the first computer science programs in the country, which happened to be in North Carolina State University… which is kind of ironic cuz it sort of mirrors my story a little bit , which is I’m one of the first data scientists… but it’s just by pure luck, you know, like I studied math, I studied a lot of different things. But, a lot of the first data scientists were sort of like renegade PhDs who were looking for a home.

Trac Bannon: 

Renegade PhDs looking for a home? I don’t know if I’d pick the term renegade to describe Rebecca. Determined, unique, non-conforming would be my first picks. She is good with math though in her words, not as good compared to the rest of her family. Her younger brother has a doctorate in physics. Academics are a curious bunch; there is often more competition than you would think.

Rebecca Bilbro: 

I think that my family would say I was always the one who was good at language, so they were more technologists and scientists, and I was the one who was good at describing things and explaining things and winning arguments.

Trac Bannon: 

Language is Dr. Bilbro’s superpower. She attended magnet schools in North Carolina. US News and World Report released data showing that North Carolina has some of the best magnet schools in the US and they start these kids in elementary school. 

While hailing from a family of engineers, Rebecca was drawn to languages, especially French. By the time she was in high school, she was immersed in Advanced placement coursework and had taken many trips to France. This was a kid that like me, loved to read the dictionary. Her passion is translation. 

Rebecca Bilbro: 

Translation is really interesting. it is a math problem because it’s kind of about learning a representation of language, like your own large language model in your mind. 

Trac Bannon: 

You are listening to Real Technologists. I’m your host, Trac Bannon, coming to you from Camp Hill, Pennsylvania. Each week we choose a unique guest behind leading Edge Tech innovation to explore their genuine stories, their true journeys. Technology touches nearly every aspect of our lives. It’s being driven by diverse perspectives and experiences of real humans.

You’re in the right spot to hear about the real technologists reshaping our world. Stay tuned for stories that will give you something to noodle on.

This dynamo is much more well-rounded than her resume and her self description would afford. When she began to look at college majors for her undergrad, she was not sure what she wanted to do. Graduating at the grand old age of 17, she applied to a number of colleges. How did she pick her alma mater, Skidmore College in Upstate New York? Based on the distance from home. 

Rebecca Bilbro: 

I applied to a bunch of schools, and of the ones where I got accepted, I picked the one that was furthest away from home. I had this notion that I needed to get far away in order to become myself, I guess. So that was not a very sophisticated algorithm, but it was like mathematically, very easy to compute.

Trac Bannon: 

Skidmore did not require freshmen to declare a major their freshman year. She naturally took French and at the same time gravitated towards math faculty because the faculty felt comfortable. It felt homey. That daughter of academics already knew how to learn math. 

She loved language and she loved mathematics. By her sophomore year, she decided to double major in English and math. That was a very unusual mix… the seeds of which were planted when she was growing up. 

As a society, we must emphasize the importance of reading and of littering the house with books, even if on a rotation from the local public library. The reason is the lifelong impact a single book can have. For me, it was the Newbery honor winner titled “My Side of the Mountain” and a protagonist named Sam Gribble. Rebecca Bilbro, the child of two college professors, would pick up a book that sowed seeds that she is still cultivating.

Rebecca Bilbro: 

So I read a book when I was a kid that I think maybe it had something to do with it. It’s not a kids’ book. I read CP Snows, the Two Cultures. It’s a lecture that he gave in the fifties, I think, it’s about his experience as a practicing academic scientist who is also interested in literature and writing.

And basically about the traumatic experience of trying to navigate those two worlds that were so inhospitable to each other… even at that time, I think reading it drew my attention to the problem and really, focused me and I’ve been sort of focused on that for my entire life since I read that book. 

Trac Bannon: 

Forever balancing the yin and yen of the liberal arts and the sciences, she continued to keep a foot in both worlds throughout her life. One of the more memorable and perhaps still frustrating teaching moments for this future world-class data scientist was made by a college English teacher. 

Rebecca Bilbro: 

I had one English teacher who kept giving me bad grades on papers and I was furious… because in my head thinking like … I’m supposed to be good at this, but she was giving me bad grades on my papers and I went in to have a one-on-one conference with her and she said… the way that you’re ordering the information is not natural from the reader perspective.

 You need to rethink about how you’re breaking your arguments down. And she said, you’re a math major, right? And I said, yes. And she’s like, when you have to do a proof, how do you do that? Her name was Phyllis Roth. 

Trac Bannon: 

Eventually it clicked for Rebecca. She needed to have a methodology for writing in the same way she did for solving mathematical theorems. I could hear in her voice and see on her face that even to this day, she’s a little irked by those initial low scores. It was a class during her senior year of her undergraduate work that would be the next catalyst and a way to glue the two sides of her left brain, right brain universe together.

Rebecca Bilbro: 

In my senior year of college, I took my first computer science class. And it was like the most amazing thing that I had ever thought about. The final project was we had to write a competitive tic-tac-toe game, and we had to build a GUI for it from scratch, and it was so hard for me.

It was magical and I worked so hard on it , like an unreasonable amount, try to catch edge cases and stuff like that. And it was my senior year and it was time to graduate. And it was too late. 

Trac Bannon: 

It was not too late though it may have felt so at the time. She has many fun quirks and habits that are endearing and that give a glimpse into the mind of this academic. Throughout her education, she has kept all of her math books. 

Rebecca Bilbro: 

I’ve actually kept all of the math books that I’ve ever had in my life because if I’m trying to remember the intermediate value theorem, I can picture that it’s like midway down on the page, like halfway through the book of this yellow book that I’ve had for many, many years now. And I have this, I guess, paranoia that if I lose the books, I’ll forget. 

Trac Bannon: 

During the lockdowns, I spent some time actually tossing out some of my high school and college books and notebooks… and yes, I may have kept a few computer programs printed out on 14 inch wide green bar, continuous computer paper. For Dr. Bilbro, the lockdowns had a different impact. After leaving North Carolina at 17, she never really returned beyond brief visits. When the worldwide lockdown started in 2020, she moved back to North Carolina to be closer to her mother, as did her brother. They relocated their families and were all together for the first time since childhood. It is now an extended reunion with their children growing up together. 

One has to wonder what the youngest generations of this academic lineage will want to study. Will they all go on to have PhDs? 

When Rebecca was prepping to graduate from Skidmore, she knew there was only one path ahead of her: continued academic study.

Rebecca Bilbro wanted to fix the CP Snow problem. She applied to grad school for both mathematics and English. When she got acceptances in both areas of study from the University of Illinois, Urbana Champagne, she dove in. She would blend both her worlds with her dissertation. Rebecca’s research was to uncover the ways that engineers communicate and the nuanced word choices, demonstrating expertise in the field.

Rebecca Bilbro: You can tell this now, if you read a paper that’s generated by an AI, the words are a little bit wrong. Like they’re close, but especially if it’s a technical paper… like technical words are so specific and so precise, and they just have these very specific words.

And so when people use words, even a little bit different or like a word that’s technically a synonym, but semantically, it’s different. You know, right away if you know that topic.

Trac Bannon: 

Urbana would be good to Rebecca. It was there that a poor graduate student looking for cheap housing moved into a group house and rented a room. It was 2005 and she would meet her future husband.

When she was close to the end of her PhD program, she applied to live in France for a year and study engineering communications in other cultures. She taught English on the side to pay the bills. France would provide another catalyst moment.

Rebecca Bilbro: 

By going to France was sort of an interrupter to that, and it made me rethink the course of my life and rethink maybe what I was supposed to do.

Trac Bannon: 

Dr. Bilbro realized that the incentives for academicians were often tied to single author publications. At some point it clicked and she realized it didn’t feel right to her. Science to Rebecca is collaborative. It was the late 2000s and it was an interesting time in the US political landscape. Perhaps it was time for a change of focus. Could she make the country better? She applied and was accepted to be a presidential management fellow. Her first stop was the US Department of Labor; it was a stop that fit her family narrative.

Rebecca Bilbro: 

You know My mother’s father was in the pipe fitters union in Baltimore. So he was a blue collar worker and my mother’s family basically was a big, family. And I think that she attributes a lot of the fact that she was able to go to school at Case Western to the pipe fitters union. I think I always had this in the back of my mind that like the Department of Labor had played some role in that. 

Trac Bannon: 

Rebecca grew rapidly at the Department of Labor like a summer tomato in miracle grow. It was Dr. David Michaels, who encouraged her to push the envelope on analytics and figure out a way to pull in the many data sets, analyze it, and prevent injury and death of the American workers. 

Rebecca Bilbro: 

I was given permission to like, install stuff on my government computer, like Python and stuff that, like nobody else was allowed to do. So because of that, I had like this sort of social capital and this encouragement and so I started building models trying to build machine learning models and that’s kind of how I ended up here.

Trac Bannon: 

Rebecca is one of those energetic people who always seemed to have multiple focus areas in parallel. While working with the Department of Labor, she was invited to be an adjunct professor at Georgetown University for continuing education certificate in analytics… one where she had recently been a student and a teaching assistant. End to end, Dr. Bilbro invested 16 years into her formal education and would go back in a heartbeat if that were a sustainable model.

Our lives are punctuated with Catalyst Moments. For Rebecca Bilbro, another catalyst moment was a government shutdown while she was with the Department of Labor. Two weeks of downtime, gave space for deeper introspection. It was time to build something; being an analyst was not enough. She’s not the first Real Technologist who had that moment of clarity.

She was drawn to the Department of Commerce as part of a data science startup. The goal was to provide data science as a service. Commerce Data Service targeted themselves then other agencies. The experience was invaluable but would not last. 

Rebecca Bilbro: 

It was going well and then I think people started to realize there was gonna be an administration change. The people who were kind of spearheading the program were political appointees and they started kind of preparing to leave.

 I think probably never meant to really be a long-term thing.

Trac Bannon: 

It was time to jump to the commercial world and begin a 2 year puddle hop cycle we see common in today’s technologists. Rebecca’s first stop was as Lead Data Scientist for ByteCubed in Virginia. She was building… finally building. ByteCubed produced custom data solutions using supervised and unsupervised machine learning. It was her next stop at District Data Labs that would introduce her to the open source ecosystem. She noticed a commonality when scientists and educators would explain models: they would instinctively draw a visualization to explain the mechanics of the model. Why not create open source tools to facilitate selecting the learning models? She first ran the idea by the data science community at PyCon 2018 but not before collaborating with her longtime colleague, Benjamin Benford, and kick-starting an initial package to enhance SciKit-learn. 

Rebecca Bilbro: 

At the time there was a machine learning library that was becoming extremely popular, called Side Kit Learn. You know, the combination of like Hadoop and Side Kit Learn becoming available, as open source projects around the same time changed the course of machine learning forever and data science forever. It’s the reason why data science probably exists.

All of the machine learning algorithms could kind of be summarized as either estimators or transformers. So transformers are things that take data, make some kind of change, project it into a different space, or clean it in some way… like normalize it, and then give back a new version of the data that’s a transformer. Or an estimator, which is something that takes data and then learns a generalization of it… learns a representation of the training data and then can be used for future predictions.

And so we thought, we can wrap visualization into that API because that’s how people are gonna be doing machine learning anyway. We’ll create a library that you can inject into that workflow with just one or two more lines of code that are really just wrapping lots of map plot lib, Python code. 

Trac Bannon: 

This was the birth of the popular Open Source Project, YelloBrick. During her time with District Data Labs, she grew as an Open Source contributor and an educator. After four years with District Data Labs and still being a little close to government, Rebecca made a dramatic shift to ICX media as their chief data scientist. Her focus was social media analytics. This shift aligned to Dr. Bilbro’s love of language, large language models, text data, and analytics. 

Rebecca Bilbro: 

I decided I wanted to try to do something that was like not government related at all, which is what took me to ICX. And basically did like social media analytics for them.

And it was a lot of text data, which is kind of my jam. 

Trac Bannon: 

She is always a prolific learner and educator. 18 months into her ICX media role, she hopped to Unisys as a Machine Learning Engineer. It was an odd time to shift jobs, though many technologists did the same… It was June of 2020 and the first peak of the universal global lockdowns. With her colleague Benjamin Bengfort as well as the dean of their Georgetown certificate program, Edwin Schmeer, a turning point decision was made, it was time to strike out on their own and found their own company: Rotational Labs and build a streaming eventing platform called Ensign. 

Rebecca Bilbro: 

It’s a streaming event platform.

And the idea is that it’s a managed version of something like Kafka. If you could build Kafka from scratch today with the last 10 or 15 years of knowledge. Kafka’s been around for a long time. It’s excellent. It’s a extremely powerful tool for this like kind of shared log abstraction.

 There was nothing else like it, there’s a lot of competitors for Kafka now, including confluent, which is selling managed kafka and all of those tools are good. They all are really focused on throughput. So they’re sort of assuming that the fundamental problem is that you’re not getting your data moved around fast enough.

And our premise is a little bit different. You really are struggling to get your models deployed, and so we need to figure out how to move data into and out of your models in a stream. So it just fits more naturally with the rest of the organization. 

Trac Bannon: 

This is Dr. Bilbro’s focus today though she continues to be ever the educator on the Data Science Faculty at Georgetown Center for Continuing Education and Professional Studies. Her mentoring and sponsoring approach is a bit different. This academician and child of Professors has collected a broad network of hundreds. She’s always reachable to give guidance, talk through a new situation or provide a letter of recommendation. She does not discriminate or focus on any one demographic though Rebecca does seek to be a role model for other women. 

Rebecca Bilbro: 

I think that there aren’t as many women in technology as would be kind of statistically kind of expected. so potentially, women, students, and other women out there who are kind of doing this work are looking for kind of a model of what they might be when they grow up.

Trac Bannon: 

Rebecca’s new corporation is her focus though she’s always looking for a parallel effort to kickstart. When I ask her directly what advice she would give to her younger self, she went silent and really contemplated it. Her initial response was curious. 

Rebecca Bilbro: 

I don’t know. I don’t know if I have a good answer to that one.

Trac Bannon: 

Thinking across our nearly two hours of conversation, there did seem to be one regret or at least acknowledgement of a choice she would make differently if she traveled back in time… finding programming earlier…

Rebecca Bilbro: 

It was too late… but if I’d done things… there’s an alternate timeline where I took that class my freshman year, and I maybe would’ve found this a little bit earlier, but I found it.

Trac Bannon: 

Rebecca has a curious and delightful philosophy for rationalizing and dealing with decisions, opportunities, and even regrets. Perhaps she learned this from Amy Pohler’s book, Yes Please, Rebecca lives in the moment and takes mental notes so she can revisit the memory or project it into the future… in essence, a bit of memory driven virtual time travel.

Rebecca Bilbro: 

I’m sort of in dialogue with many versions of myself across time. So maybe I don’t have any specific advice for that version of me, but I’m in touch with her.

And I remember what it feels like to be her. And I also remember what it feels like to be the me in 30 years, and I I go and visit with them sometimes and check in when I’m going through something that feels challenging in the present moment, or when I’m remembering a challenge from a different moment in time, you know, kind of go and have a little chat and check in.

Trac Bannon: 

And that’s a wrap for today’s episode of Real Technologists. I want to thank my guest, Rebecca Bilbro for sharing her story. Your insights and experiences are truly inspiring. I’m grateful for the opportunity to share them with the audience. This podcast is a Sourced Network production and updates are available weekly on your favorite audio streaming platform. Just search for real technologists and consider subscribing. Special thanks to our executive producer, Mark Miller, for making this show possible. Our editor and sound engineer, Pokie Huang has done an amazing job bringing this story to life. Thank you both. The music for today’s episode was provided by Blue Dot Sessions, and we use Descript for spoken text editing and audacity for the soundscaping. The show distribution platform is provided by CaptivateFM making it easy for our listeners to find and enjoy the show. 

That’s all for today, folks. This is Trac Bannon. Don’t forget to tune in next week for another intriguing episode of Real Technologists and something new to noodle on.