I am a CS PhD student at Stanford, co-advised by
Anshul Kundaje and
Brian Trippe, working on
generative models for biomedicine. Previously, I was an ML engineer at
the Broad Institute in the
ML4H group and a
Senior AI Scientist at Genentech in
BRAID.
I have led machine learning projects spanning methods research and
applied work in small-molecule discovery, genomics, and clinical risk
prediction.
-
Education
PhD Computer Science — StanfordStanford, CA — 2024–Present
-
Developing novel generative modeling and finetuning methods for
biology (co-advised by Brian Trippe and Anshul Kundaje).
-
Stanford Graduate Fellow with three years of fellowship funding.
-
Coursework in information theory, measure theory, SDEs, game
theory.
BS Computer Science and Math — Harvey Mudd CollegeClaremont, CA — 2015–2019
-
GPA: 3.87, dean's list all eight semesters, graduated with high
distinction and math departmental honors.
-
Henry A. Krieger Prize for Outstanding Promise in the Field of
Probability, Statistics, or Operations Research (1–2 per
graduating class).
-
Experience
Senior AI Scientist — Genentech Early ResearchSouth SF, CA — 2022–2025
-
Developed and implemented (in PyTorch) graph deep learning,
semi-supervised learning, uncertainty quantification, and
generative modeling methods for small molecule, macrocyclic
peptide, single cell RNA-seq, and regulatory genomic applications.
-
Received gRED Recognition Award (top 10% research employee) both
eligible years.
-
10xed binding strength of selected peptides over expert selection
using novel deep learning statistical inference method.
-
Designed small molecule selection algorithm responsible for 100s
of synthesized molecules and several novel candidates.
ML Engineer II — Broad Institute Machine Learning for Health
GroupCambridge, MA — 2019–2022
-
Collaborated with clinical researchers on deep learning
cardiovascular imaging and genetics applications resulting in two
patents and improved disease understanding.
-
Co-wrote open-source TensorFlow repo for multi-modal, multi-task
clinical deep learning still used for all of the team's projects.
Research Affiliate (part-time) — MIT Computational Cardiovascular
Research GroupCambridge, MA — 2019–2020
-
Developed and published ECG foundation model resulting in several
clinical and machine learning follow-up works.
ML Engineering Intern — AdRollSF, CA — Summer 2018
-
Implemented inverse transform sampler of statistical model of
advertising conversions used in customer dashboard.
ML Engineering Intern — YelpSF, CA — Summer 2017
-
Implemented spam photo filtering backend with millions of uses per
day using Amazon Redshift and learned industry standard coding
practices.
-
Selected Publications
Calibrating Generative ModelsarXiv (under review), 2025
Smith, H., Diamant, N., Trippe, B.
-
Contribution: Extended method and ran experiments for LLMs and
discrete diffusion models.
A cell atlas foundation model for scalable search of similar human
cellsNature, 2025
Heimberg, G., Kuo, T., DePianto, D., Salem, O., Heigl, T., ...,
Diamant, N., et al.
-
Contribution: Designed and implemented interpretability method.
Conformalized deep splines for optimal and efficient prediction
setsAISTATS, 2024
Diamant, N., Hajiramezanali, E., Biancalani, T., Scalia, G.
- Contribution: Invented method and ran all experiments.
Improving graph generation by restricting graph bandwidthICML, 2023
Diamant, N., Tseng, A., Chuang, K., Biancalani, T., Scalia, G.
- Contribution: Invented method and ran all experiments.
Patient contrastive learning: A performant, expressive, and
practical approach to electrocardiogram modelingPLoS Comp Bio, 2022
Diamant, N., Reinertsen, E., Song, S., Aguirre, A., Stultz, C.,
et al.
- Contribution: Invented method and ran all experiments.
Last updated: April 2, 2026