Gerstein Lab, Yale University — Research Assistant
New Haven, CT
- Worked on a multi-modal RNA, DNA, protein, and small molecule bio LLM reasoning model. Designed multimodal language model for tasks with small molecules.
Andrew Tran
Previously built data infrastructure and product at HRT and Ramp. Graduating in May 2026 with a B.S./M.S. in Computer Science at Yale. Interested in high-performance systems, the Python data science ecosystem, distributed ML infrastructure, and applied machine learning.
Experience
- Oct 2022 – Apr 2026
- May 2025 – Aug 2025
Hudson River Trading — Software Engineer Intern
New York, NY
- Designed a FUSE filesystem in C++ for efficiently serving source files and built Python tooling for analyzing resource usage across a distributed compute cluster.
- May 2024 – Apr 2025
Ramp — Software Engineer Intern
New York, NY
- Led development of a card delivery platform to increase usage of card payments across the product.
Show 4 more Show less
- May 2023 – Aug 2023
HealthPartners — Software Engineer Intern
Minneapolis, MN
- Developed and maintained payment flows for insurance portals using Vue.js/Nuxt and Java/Spring Boot.
- Jun 2022 – Aug 2022
Abbott Laboratories — Software Engineer Intern
St. Paul, MN
- Created GUI software for editing physical properties of heart medical devices stored in SQL databases.
- Jun 2021 – Aug 2021
Abbott Laboratories — R&D Engineer Intern
Minneapolis, MN
- Designed an algorithm for aligning time-shifted electrocardiogram data for quality control.
- Jun 2020 – Aug 2020
Abbott Laboratories — Process Development Engineer Intern
Minneapolis, MN
- Developed programs for programmable logic controllers to improve efficiency and quality for heart medical device manufacturing.
Selected Projects
-
GPU Topology-Aware PyTorch Model Partitioning Engine
Python · PyTorch · PyTorch Distributed
Designed a system to automatically partition PyTorch computation graphs based on GPU network topology.
-
Single-Cell Chemical Perturbation Hybrid Graph Neural Network
Python · PyTorch · PyTorch Geometric
Combined variational autoencoders and graph neural networks to predict differential gene expression from chemical perturbations.
-
Keyboard Video to Text
Python · HuggingFace · OpenCV
Designed a model to predict the text typed from an overhead video of typing on a keyboard.
Publications
-
MolLM: A unified language model for integrating biomedical text with 2D and 3D molecular representations
Bioinformatics, June 2024
Xiangru Tang, Andrew Tran, et al.
Education
Sep 2022 – May 2026
Yale University
B.S./M.S. in Computer Science; Certificate in Data Science
Selected coursework
Trustworthy Machine Learning, Probabilistic Machine Learning, Theory of Statistics, Deep Learning on Graphs, Natural Language Processing, Distributed Systems, Operating Systems, Computer Architecture, Computer Networks, Compilers, Algorithms.