Jesse Rusak is a member of technical staff at Open Athena’s Boston office, where he works on the ocean modeling project Samudra. In this Q&A, he digs into the challenge of modeling the ocean, the idea that a general-purpose large language model can lend its infrastructure to scientific foundation models, and the role elephant seals can play in climate research.
What was your first experience with AI technology?
Jesse Rusak: Back in my undergrad, I did some machine learning research on brain-computer interaction. So we had subjects with EEGs on their head, and based on brain activity, we were trying to predict what button they were going to push before they pushed it, using very old-school support vector machine technology. So that was really my first interaction with machine learning.
For more modern AI, back in 2015, I joined this startup called Semantic Machines, which was building chatbots for customer support and other similar things, back before ChatGPT came out. I spent a lot of time there learning about the challenges and technologies behind then-modern natural language processing (NLP) and AI.
What drew you to work at Open Athena?
After leaving Microsoft, I took some time off to try and think about what I wanted to do with my life. I basically wanted to find a place where I could work with great people and use some of the skills I had built around AI, and one that would help the Earth and the people that live on it. To me, the most obvious, urgent thing would be to somehow accelerate climate research, as a way of making use of AI, as well as trying to be helpful.
So I talked to a bunch of universities and labs, and groups that were helping researchers with their machine learning problems. After several of those conversations, I really came away pretty disappointed—or maybe hopeless is a better word—about how much I could actually help as an individual. These labs and central groups and so on, they’re underfunded, they don’t have access to the GPUs they need, they’re in these institutions that don’t value software engineering. And I would be the only software engineer in most of these cases. So I came away feeling pretty disappointed about that as an option.
And then I got connected to Jeff Hammerbacher, the founder and CEO of Open Athena. I had a great conversation with him, where he pitched me the idea of Open Athena. It was all the things I wanted to do to help academia and accelerate research and bring software engineering talent to those important problems, but done with a team, and with real funding, and with his own connections, and so on. It was a revelation, like, “Oh right, we can do this, we just need to approach this problem in a better way.” And so I was pretty sold right after that first conversation.
Which Open Athena project do you primarily work on these days?
There’s two, but the one I’ll talk about is, myself and Alex Merose have been working with labs at NYU and MIT on an ocean modeling project for about a year and a few months now. Our goal is to make it cheaper and more accurate to predict what the ocean will do over long periods of time, which is a critical ingredient in understanding what the climate will do over long periods of time: decades or centuries.
And it’s a really challenging machine learning problem because we don’t have a ton of information about the state of the ocean. All we can really tell is what’s happening at the surface. They have these buoys that float around that tell you a little bit about the interior, but we really don’t know what’s going on deep below, and we’ve only actually been observing the ocean for a few decades. We don’t have huge, long histories.
So we have a relatively small amount of very sparse data, and we want to predict for decades or centuries what’s going to happen. And so we’re trying to use machine learning to fill in those gaps and help extrapolate what will happen, so that scientists can probe why the ocean acts the way it does and how it might act in the future.
And did you have any background in ocean dynamics before you started on this project?
My first internship was actually at a fluid dynamics company, in my undergrad. They were trying to simulate fluid dynamics for military reasons, like building bunkers that could withstand shock waves and things like that. But it wasn’t really at the ocean scale.
Do you also do non-project specific work at Open Athena?
In addition to helping these labs with the particular machine learning problem that they’re interested in, I’m also co-leading an internal team. We’re trying to figure out how we can have a common infrastructure that can support all the different science problems that Open Athena is helping to work on—and hopefully others beyond that—on top of the Marin project’s deep investment in machine learning models and infrastructure. So we’re really trying to figure out: how easy is it to take varied scientific problems across different domains and frame them in such a way that we can then reuse these same models, pipelines, compute resources and so on—to make it faster to iterate on these models. And so we’re figuring out what pieces are missing, what works already, and how we can improve that workflow for ourselves internally, and for scientists out in the world.
What do you like best about working here?
The kinds of challenges that we’re facing are pretty unique. I don’t think a lot of companies are thinking about, “how do we broadly accelerate science across academia? How do we build models that can work across all these different domains?” And so I think that challenge, and its ambitiousness, is the thing that is most exciting to me.
What does a typical work day look like for you?
I’m based in the Boston office with one other co-worker. So usually I’ll come in in the morning and go through a bunch of discussions that happened the previous night with our folks on the West Coast, our grad students that we work with, who are often up late. And usually I’m preparing for a couple of group meetings and stuff during the day. I then try to have a big block of focus time somewhere in there, to iterate on a particular model or do some design, often through talking with either a person or a large language model as a partner in that process. And then I often have a bunch of code review and other things like that to do. So much of our job now is review and not writing code in the first place, so that tends to take up a good chunk of my afternoons.
Do you have a favorite piece of science trivia?
Maybe I’ll relay one that we had from a PI, Laure Zanna, who came to talk to us at our offsite recently. She shared that one of the hardest places to understand what the ocean is doing is under the sea ice, because you really can’t look down underneath that with satellites or other up-in-the-air strategies. And so they put monitors on top of the heads of elephant seals, who then dive below the ice, in order to get a sense of what’s happening down there—temperature, salinity, and other things. That tells you just how hard it is to understand what is happening in the ocean, that we have to rely on our seal friends to do scouting for us.
What’s a fun hobby or activity you like to do outside work hours?
I actually just finished up a game side project that I had been working on with my brother for the last two and a half years, just about a month ago, which we had a lot of fun working on. Though I’m also glad that it’s finished! It’s a fun little thing for Panic’s Playdate system, which is a sort of retro console that’s pretty fun to build little games for. The game is a space-themed roguelike, where you’re this tiny ship that has a huge beam, and you fight enemies all over different planets in space. It’s called Beampunk.
Last but not least, what’s your signature reaction emoji?
I am a big fan of using the bow emoji 🙇 to say thank you. I have been trying to get that to catch on more within the various emoji spaces I inhabit.
Cite this post
@misc{bushwick2026_meet_our_team_jesse_rusak,
author = {Bushwick, Sophie},
title = {Meet our team: Jesse Rusak},
year = {2026},
month = {sep},
howpublished = {\url{https://www.openathena.ai/blog/meet-our-team-jesse-rusak/}},
note = {Open Athena Blog}
}