Protein AI
ProtDiffusion
A PyTorch implementation exploring diffusion-based protein structure prediction.
PyTorchdiffusionprotein structure
Project notes ↗Selected work
Projects
Work organized by the capabilities it develops: scientific modeling, modern ML engineering, and systems fundamentals.
Scientific ML
Scientific ML
AlphaFold2-mini
An implementation of the AlphaFold2 architecture from first principles: Evoformer, MSA row/column attention, outer product mean, triangle attention and multiplication, Invariant Point Attention, structure module.
attentionprotein foldingtransformers
Project notes ↗Biological foundation models
scGPT-mini
An implementation study of scGPT for single-cell transcriptomics.
transformerssingle-cellomics
Project notes ↗Computational biology
Breast cancer prognosis
A machine-learning project for prognosis prediction from transcriptomic and clinical data.
RNA-seqclinical dataML
Project notes ↗AI Engineering
Agentic AI
BioAgent
A biology-focused tool-calling agent fine-tuned with SFT/LoRA on open-source language models.
tool callingSFT / LoRALLMs
Project notes ↗AI Engineering
ML systems & inference
Implementation-led study of the systems choices behind useful, reliable LLM applications.
inferenceservingreliability
Project notes ↗Systems Engineering
Database internals
MiniKDB
An in-memory persistent embedded key-value database.
C++key-value storepersistence
Project notes ↗Database internals
FensterSQL
A relational database system with an emphasis on SQL and query optimization.
CSQLquery optimization
Project notes ↗