How do you build a career in one of the most promising corners of tech?
Amy Heineike’s answer might not be what you’d expect. At this point in her career, she’s VP of Engineering (EMEA) and one of the original team members at Primer, a company at the forefront of the NLP revolution. But she didn’t exactly take a direct path to get there.
As a kid growing up in the small British town of Derby, Amy was interested in everything. She liked making things. She liked puzzles. She liked English as much as math and science. Then at university, she took a nonlinear dynamics class and remembers having her brain “slightly blown” by complexity theory.
But when she graduated with a degree in mathematics, she still wasn’t exactly sure what would come next. “Data science” wasn’t a field that existed yet. “I remember thinking a lot about doing different master’s programs,” she says. “And my wonderful father said to me, ‘Get a job. Get a job. Get a job.’”
She got a job, but it didn’t exactly lead to immediate fulfillment. She was working on transportation planning, and “it was fine, but I was a bit bored.”
That early-career malaise created the spark Amy needed to leapfrog out of her boring first job and onto a path that would lead her to NLP.
In our first episode of Technical Women, Amy talks to hosts Natalie Vais and Renee Shah about how her winding path led her to build completely new language processing models. Subscribe to the Technical Women newsletter and to your podcast feed and listen in!
When she was working at that boring first job, Amy started reading voraciously. “I knew I didn’t get it right the first time, and I wanted to find something I was interested in,” she says. She picked up a book on complexity economics and was hooked.
“I looked up the author and realized he also lived in London, and he had a consultancy, and I persuaded myself that they probably needed to hire somebody.” So she wrote to the firm and applied for a job. They didn’t have an open position, but she eventually got a job there anyway, and she was off to the races toward work that kept her attention.
“I’m pretty curiosity-driven,” she says. “That gave me a chance to open a door to something that turned out to be really cool, working on some super interesting problems. Sometimes I talk to people who are early in their career, and they’re trying to work out the right move to make. And I’ve learned that it’s okay if opportunities end up being a bit wrong. Because if you’re bored, you can spend time figuring out what’s exciting. And then maybe you’ll find the door, and you can jump through it.”
Her advice: “Things are changing quickly, and it’s hard to know what the future holds. It’s okay if you don’t know how your career is going to end up. You just learn what you can from each step.”
Finding other thinkers and tinkerers
It should be no surprise that Amy is curious about how to best build a team. When she started hiring a team in the UK, she realized quickly that if she just looked for people who fit the traditional pattern — computer science degree from a top university, followed by an internship in Silicon Valley — she was really limiting her options.
Instead, she looks for “people with computational science backgrounds, who started as software engineers, but got curious about what’s possible with machine learning, or people who retrain themselves — maybe they were a journalist to start with and then they went into a boot camp and picked up all the skills to broaden themselves.”
“So when we go into interviews, we’re not narrowly looking for whether they know and can execute on one technique. We want to find out if, given some kind of problem, they have a nice way of breaking it down.”
“I think when you kind of have that philosophy to how you hire, you naturally end up picking up people who’ve come through some very different paths to get into the job. And when you’re open to people coming through different paths, and you recognize different skill sets and value them, you end up making it a lot easier to hire people who are quite diverse. Who match different patterns.”
How does Amy apply her curious spirit to her work in NLP? She has been working in NLP for 7 years, on the team building Primer’s core code and applications. That means she has seen the models completely evolve. And one of the core challenges is still figuring out how to use data to draw interesting new conclusions. At this point, a lot of people realize NLP is out there, she says, “but it’s still quite hard to figure out how to make it useful.”
Part of the challenge lies in the kinds of data NLP is processing. We’re used to search-based technology, she says. We know how to Google a question and find an answer. But how do we gather collective knowledge and nuanced cues to understand broader conversations and trends?
Take Twitter, she says: “My Twitter feed is full of really interesting people tweeting. If I want to find a tweet from somebody, I can look them up. But what if I want to say, ‘On aggregate in my Twitter feed, are they super worried about COVID at the moment, or are they not that worried and they’re feeling optimistic? What are the big trends people are focused on?’”
Right now, she says, she just has to doom scroll for a while and synthesize the overall trends in her head. But Primer is building tools that could do that synthesis for her — that can read and write like an analyst.
Amy feels excited about what’s coming next from NLP. “What’s fun about working with language is it’s so rich. People can express so many wonderful ideas with language. But it takes a bit of an imagination leap to work out [how to make models that process all that language].”
“I’m always asking, ‘What else could we learn from this?’ It’s endlessly interesting.”
Subscribe to Technical Women wherever you listen to podcasts and directly to our newsletter to get next episodes sent to your inbox. Who should we interview next? Email us at firstname.lastname@example.org.