Home on Victoria Dyster, PhDhttp://victoriadyster.com/Recent content in Home on Victoria Dyster, PhDHugo -- gohugo.ioen© 2026 Victoria Dyster, PhDSun, 14 Dec 2025 00:00:00 +0000From CSVs to Iceberg: Scaling a Genomics ETL Pipeline for ML Training on a budgethttp://victoriadyster.com/projects/csv-to-iceberg-methylation-analytics/Sun, 14 Dec 2025 00:00:00 +0000http://victoriadyster.com/projects/csv-to-iceberg-methylation-analytics/How replacing a CSV-join pipeline with Apache Iceberg and a long-format data model cut an ETL pipeline from ~15 minutes to a minuteBuilding a Private MLOps Platform on AWShttp://victoriadyster.com/projects/private-mlops-platform-aws/Mon, 01 Sep 2025 00:00:00 +0000http://victoriadyster.com/projects/private-mlops-platform-aws/How I deployed MLflow as a authenticated experiment tracking server on AWS and integrated it into a reusable ML toolkit.CI/CD for a Genomic Data Pipeline: Testing, Security, and Multi-Environment Deploymenthttp://victoriadyster.com/projects/ci-cd-genomic-data-pipeline/Mon, 23 Jun 2025 00:00:00 +0000http://victoriadyster.com/projects/ci-cd-genomic-data-pipeline/How I built a CI/CD system for a Nextflow methylation sequencing pipeline: from pre-commit linting through four layers of testing to automated promotion across development, staging, and production, all backed by reusable GitHub Actions and container image promotion via ECR.Using Learning Curves to Know Whether More Data Will Helphttp://victoriadyster.com/blog/learning-curves-sample-size/Mon, 10 Mar 2025 00:00:00 +0000http://victoriadyster.com/blog/learning-curves-sample-size/Learning curves won’t tell you exactly how many samples to collect, but they will tell you whether collecting more is worth it at all. In domains where each sample costs real money, that’s the question that actually matters.Building a Reusable ML Toolkit for Genomic Modelshttp://victoriadyster.com/projects/ml-toolkit/Sat, 01 Mar 2025 00:00:00 +0000http://victoriadyster.com/projects/ml-toolkit/A Python package with a YAML-driven pipeline builder and a prediction CLI.When Random Features Work Just as Wellhttp://victoriadyster.com/blog/when-random-features-work-just-as-well/Mon, 20 Jan 2025 00:00:00 +0000http://victoriadyster.com/blog/when-random-features-work-just-as-well/On the counterintuitive finding that randomly selecting features from high-dimensional genomic data often matches the performance of careful feature engineering and why that makes mathematical sense.Scaling ML Training for Epigenetic Age Predictionhttp://victoriadyster.com/projects/epigenetic-age-prediction/Fri, 15 Nov 2024 00:00:00 +0000http://victoriadyster.com/projects/epigenetic-age-prediction/How parallelising hyperparameter tuning on SageMaker turned a single-instance grid search into a 100x faster training workflow.Building a Self-Service Analysis Environment for Data Scientistshttp://victoriadyster.com/projects/self-service-ec2-platform/Fri, 15 Mar 2024 00:00:00 +0000http://victoriadyster.com/projects/self-service-ec2-platform/Designing and building a Python CLI that lets data scientists create, manage, and safely shut down cloud research environments without needing to know Terraform or the AWS console.Palidis: A C++ Algorithm for Discovering Insertion Sequences in Metagenomic Datahttp://victoriadyster.com/projects/palidis-insertion-sequence-discovery/Tue, 10 Oct 2023 00:00:00 +0000http://victoriadyster.com/projects/palidis-insertion-sequence-discovery/How I built a maximal exact matching algorithm in C++ with two-bit encoding to discover novel mobile genetic elements from metagenomic sequencing data and how it found applications from antimicrobial resistance surveillance to gene therapy manufacturing.Glorified Excel: The Dashboard Feature Request You Should Push Back Onhttp://victoriadyster.com/blog/glorified-excel/Fri, 01 Sep 2023 00:00:00 +0000http://victoriadyster.com/blog/glorified-excel/In biotech, the most common dashboard feature request is a filterable, sortable table. The fastest solution is usually a CSV download and the spreadsheet software people already know.FormulAI: Using LangChain to Generate Skincare Formulationshttp://victoriadyster.com/blog/formulai-langchain-formulations/Sun, 11 Jun 2023 00:00:00 +0000http://victoriadyster.com/blog/formulai-langchain-formulations/A hackathon project that chains LLM calls with a product ingredient database and Wikipedia to generate skincare formulations — ingredients, assembly protocols, and allergen warnings.How to Impress Someone with a Bioinformatics Pipelinehttp://victoriadyster.com/blog/how-to-impress-someone-with-a-bioinformatics-pipeline/Sun, 02 Apr 2023 00:00:00 +0000http://victoriadyster.com/blog/how-to-impress-someone-with-a-bioinformatics-pipeline/What makes a good bioinformatics pipeline? A short, non-technical take on the things that matter — from user requirements to reproducibility to knowing when good enough is good enough.