Skip to main contentA logo with &quat;the muse&quat; in dark blue text.
Advice / Career Paths / Career Stories

Here’s What it’s Like to Work on Ground-Breaking AI Projects at Meta

Angela F., a research scientist at Meta
Angela F., a research scientist at Meta.
| Courtesy of Meta

If you’d asked Angela F. as a college student if she thought she’d one day work on groundbreaking artificial intelligence (AI) projects, her answer would have been a resounding “no.”

“I was pre-med because I was passionate about helping people,” she says. “I didn’t switch to consider a career in tech until the latter half of college.”

But once she discovered data science, Angela was hooked. She pivoted immediately and began pursuing a future in tech. Her first stop? Meta, where she participated in the Meta Immersion Program, which teaches people without a data science background to code.

“When I discovered the Immersion Program, I knew it was the place for me,” she says. “Even though I came in with a nontraditional background, I got to learn how to code at a large company with complex frameworks, libraries, and codebases. I realized being different is a strength. I don't need to be a code machine. I can be someone who pursues interesting problems across areas that help me learn and grow.”

Today, Angela—who recently relocated from Meta’s Paris office to New York—is doing just that as a research scientist at Meta. Her work focuses on generative AI and large language models (LLMs), and she also collaborates closely with the machine learning and product teams.

Here, Angela shares the most meaningful AI project she’s worked on, how Meta’s culture inspires her to be her true self, and the one skill that has helped her succeed.

What are you responsible for as a research scientist?

Currently, I work on Meta’s Llama LLM efforts, including Llama 2, which we released in July 2023. I also worked on the Meta line of generative AI products. Day to day, I do a little bit of coding, some experimentation, and a lot of testing of the product and our systems broadly. I also have discussions about what research directions we should be pursuing. And when I have time, I read industry papers to stay up to date on what’s going on.

What do you like most about working in research? And what is the biggest challenge?

I like all of the things you can learn and the different problems you can work on. I’ve had a wonderful experience working on everything from on-device AI to machine translation to summarization to writing stories for children to the Meta AI Research SuperCluster and now language modeling. Being able to apply your skills to this wide variety of problems is super interesting.

The challenge, of course, is that research is about discovering new things, and that’s pretty hard! You need to have a lot of patience and be willing to try many things many times since most of your ideas turn out not to work.

Through your research in AI, you made significant contributions to No Language Left Behind. Why was working on this LLM project so meaningful for you?

No Language Left Behind is very meaningful to me personally as well as many others on the team who speak low-resource languages that are not supported well by existing translation technologies. My family speaks a local Chinese dialect, and there’s no translation service for it. It’s not taught in school anymore, so my kids might be the last generation to speak it.

What started as a passion project grew into a major interdisciplinary initiative, where we worked with ethicists, linguists, and sociologists. It was a privilege to see how all of these different domains of research came together in one project. We partnered with Wikipedia to make this available for many languages on the platform that didn’t have translation systems before, so editors writing new articles can have a starting point. It was the first time I saw research production come to life.

How has Meta encouraged and supported your career growth and development?

There are many opportunities available within the company to support your personal interests and desire for growth and learning. For example, I transitioned from working as a data scientist into research engineering and then became a research scientist. I’ve been able to work on both research and products. Being able to follow your interests and passions is one of the most critical parts of career growth. That’s one of the main reasons why I love my job here.

How has the culture at Meta impacted you personally?

Earlier in my career, I struggled to show up authentically at work. I didn’t feel comfortable expressing my personality, decorating my desk, or sharing my hobbies, but the inclusive culture at Meta helped me accept who I am. I put keepsakes on my desk and all my team members know about my handbag obsession.

Why is now an especially exciting time to join Meta, with regards to its AI-related projects?

Meta has a great culture around research and, in particular, the open sourcing of research work, which drives the whole community forward. It couples this with immense potential for product applications, meaning the work can really extend all the way. In particular, I’m excited about the possibility for generative AI applications across the wide variety of apps that Meta creates.

What is the most important skill to have to succeed in an AI career?

The most important skill is passion and a desire to learn and grow. Especially if you are transitioning into AI, it can be intimidating because it’s a very complex space that’s evolving everyday. It’s super critical to embrace this and give yourself time to learn and ramp up.

You recently relocated from Paris to New York. What do you like most and least about your new home?

My least favorite thing is that a croissant costs like $6 in New York and it’s not even good! My favorite thing is definitely being closer to my family, especially after the pandemic, and working again with some colleagues in the U.S. who I didn’t work with as closely while I was in France.