ZTF Faces: Summer students and their mentors
Harlan Phillips (University of California, Berkeley)
My name is Harlan Phillips, I began studying at Bakersfield College, majoring in Computer Science.
I had the incredible opportunity to intern at NASA Ames Research Center, which fueled my passion for
space exploration. Now, as an Electrical Engineering and Computer Science major at UC Berkeley,
my fascination with space continues to grow. A pivotal moment for me was attending a conference
where I heard an astronaut speak about the 'Overview Effect'— a profound psychological shift
astronauts experience in space. That experience deepened my interest in space and astronomy,
inspiring me to pursue further learning and research. Outside of academics, I enjoy
snowboarding, boxing, reading, and mentoring youth.
Anna Ho (Cornell University)
I am an assistant professor of astronomy at Cornell, studying cosmic explosions such as gamma-ray bursts and
supernovae. Outside research, I sing in a local community choir and volunteer at the Lab of Ornithology.
Creating the largest catalog of cosmic explosions
This project will enable the creation of the largest catalog of extragalactic
transients to date. Current catalogs cap in the tens of thousands, but this project will expand that number
by an order of magnitude, providing access to over a hundred thousand transients. This will offer
unprecedented opportunities for researchers to perform detailed statistical analyses on the rates
and types of transients, increasing the potential for new discoveries.
The web application plays a critical role, as it integrates data from major sky surveys: Pan-STARRS, Legacy Survey, VLASS, and ZTF (via Caltech’s Kowalski API). This allows the application to provide up-to-date, comprehensive information. With features such as interactable WISE Plots, Centroid Plots of nearby sources, and Photometry Plots, researchers can deeply engage with the data, visualizing key metrics in real-time and exploring trends in a user-friendly interface. This powerful tool is designed to be accessible to over 100 users, giving access to high-quality transient data and paving the way for more efficient, large-scale studies.
At its core, the project also involves a machine learning model that integrates tabular, photometric, and image data. As more classifications are performed, the model continuously improves, eventually automating the classification process once it reaches 95% accuracy. This automation will further accelerate discoveries and deepen our understanding of the universe.
How do you pick a project to work on?
I choose projects based on how unfamiliar or challenging they are to me. I prefer stepping out of my comfort zone because tackling new problems tests my problem-solving skills and increases the sense of satisfaction when I complete them. At the same time, I look for projects that spark my interest—areas I've always wanted to explore—where I can apply my existing knowledge while learning something new.
How do you formulate a project for a student?
It depends on the student. Generally, I look for projects that align well with a student's skills, goals, and interests (and vice versa), that have the potential to lead to an impactful result, and that have flexibility (e.g., don't absolutely need to be done on a very short timescale).
What is the number one quality you look at when you select a mentor to work with?
The number one quality I look for in a mentor is the passion to uplift others. I strongly believe that those who give the most receive the most, and mentors who genuinely want to help others grow are the ones I admire and aspire to be like. Another important quality is patience—understanding that they were once in my shoes and guiding with that empathy in mind.
What is the number one quality you look at when you select a student to work with?
Each student has different strengths, so it's hard to answer this question! Maybe I'll say: an ability to work well with others, since our group collaborates on transient science.
Was there a specific moment during the summer research work that was particularly exciting?
How about challenging?
One of the most exciting moments during my summer research was when the entire research group came together to submit a proposal that was due in just a few hours. It was thrilling to see everyone coordinate so efficiently, ensuring every detail was perfect under a tight deadline. On the challenging side, there were several moments during the development of the application, especially dealing with bug fixes. One instance that stands out is when I spent days creating a function to generate images from FITS data, only to realize later that my mentor had existing code I could have referenced. This taught me the importance of communicating my struggles and not hesitating to ask for help, even if I think I might be bothering someone.
Was there a specific moment during the summer research work that was particularly exciting?
How about challenging?
Harlan made impressive progress on his project very quickly, which enabled us to take it much further over the summer than we had been expecting.
What do you think is the most valuable thing you learned this summer?
The most valuable lesson I learned this summer is the importance of perseverance. I realized that with enough time and effort, any problem can be solved. There were moments when I felt like giving up because certain features seemed impossible to implement, but I always managed to figure it out. This experience taught me to stay creative and open-minded, never ruling out possibilities too quickly. If we had believed that getting to the moon was impossible, it would have never happened.
What do you think is the most valuable thing you learned this summer?
Within the context of this project, I learned a lot from Harlan about web-application development.
In science, answering one question also results in asking oneself a set of new ones. What are
these for you at the end of this project?
- What other transients have yet to be officially classified?
- What patterns might emerge from a statistical analysis of the rates of different transients?
- Will further discoveries in transients enhance our understanding of phenomena like jets?
In science, answering one question also results in asking oneself a set of new ones. What are
these for you at the end of this project?
How can we most effectively group, or classify, our sample of extragalactic transients?
The universe never fails to surprise us, but did you manage to surprise yourself this summer?
If yes, how?
I definitely surprised myself with the result of the project. Throughout the process, I was so focused on getting as much done as possible that I didn’t fully appreciate what I had achieved. When I look back at my poster and reflect on my presentation, I realize how incredible it is. When explaining my work to others, I often have to clarify that I didn’t just work on the machine learning model or produce a few plots. I built the entire codebase and web application from scratch and developed the ML model—all within just 10 weeks.