01

About Me

Logan Day

I'm finishing my MS in Computer Science at Western Washington University, where I bring a biochemistry background to computational research. I study how insertion mutations structurally impact protein interfaces, and I build machine learning systems for network intrusion detection -- most recently a contrastive learning approach for detecting DNS exfiltration attacks. I also enjoy building web tools, and keeping up with new technologies in the javascript ecosystem.

When I'm not in the lab, I'm usually playing music with friends, outside exploring nature, or reading a good book.

Location

Washington State

Focus Areas

Bioinformatics, ML, Web Dev

Degree

MSCS (GPA: 3.94)

02

Publications & Presentations

Best Presentation (Honorable Mention) 2025

Structural Impacts of Insertion Mutations on Protein-Protein Interfaces

Computational Structural Bioinformatics Workshop at ACM-BCB 2025

  • Proposed ΔFA, a novel metric for quantifying structural changes at protein-protein interaction interfaces caused by insertion mutations
  • Exhaustively generated and analyzed indel mutants using Rosetta computational modeling to study effects of interface distance, residue type, and secondary structure
In Review 2026

Adversarial DNS Exfiltration: Framework and Defense Evaluation

  • Developed ContraDNS, a contrastive learning architecture for stateless DNS anomaly detection that trains exclusively on benign traffic
  • Multi-scale character-level CNN encodes queries into embeddings with DNS-specific augmentations; detects attacks as out-of-distribution samples via distance thresholding
  • Cross-dataset evaluation revealed no single detection paradigm catches all attack types, demonstrating the need for multi-modal defense strategies

03

Personal Projects

Slab Scraper

Svelte / GraphQL

A tool that scrapes venue websites powered by VenuePilot, collects show information into a single events list, and exports the data as CSV. Built for the local online music zine The Slab.

  • Queries VenuePilot's public GraphQL API client-side to aggregate show listings across multiple venues
  • Discovers new venue IDs by scraping calendar pages via CORS proxy, with localStorage persistence
  • Date filtering and CSV export for easy use in print zine production
SvelteGraphQLLocalStorageCSV ExportCORS Proxy Try it

04

Technical Skills

Specializations

Structural BioinformaticsMachine LearningWeb Development

Languages

PythonTypescriptRustC/C++RSQLHTMLCSS

Tools & Frameworks

SvelteReactTailwind CSSShadcnDockerPyTorchScikit-learnPandasRosettaJupyter

Workflow

VimTmuxUnix tools

05

Get in Touch

Interested in collaboration, research, or just want to connect? Feel free to reach out.

Built with SvelteKit & Tailwind CSS