Portrait of Aaron Brooks

$ whoami_

Aaron Brooks, Ph.D.

Scientific leader building AI-native tools that transform and accelerate the life sciences — working at the intersection of synthetic biology, genomics, and agentic AI.

cat vision.md

# vision

In the future, we won't just evolve genomes. We'll design them.

I started my career modeling how genomes are regulated, then spent years helping build the largest synthetic genome to date: Sc2.0, a multi-chromosome synthetic yeast genome, alongside the Steinmetz, Boeke, and Cai labs.

That work taught me that genome design is still slow, imprecise, and unpredictable. Genomes are massively complex, evolved systems that carry the convoluted logic of their history, which is more than anyone can hold in their head. With AI, we can finally begin to understand them at scale, in context, and on their own terms.

But first-principles reasoning only goes so far with evolved systems. Designs have to be tested. Even a data center full of Einsteins would still need experiments to guide its thinking. I firmly believe progress in biology with AI depends on tight lab-in-the-loop integration: fast design–build–test–learn (DBTL) cycles that gather data efficiently, so every experiment sharpens the next design.

That's what I'm building toward: agentic systems that design, simulate, and test biology fast enough that building a genome from scratch stops being a moonshot and starts becoming a roadmap.

cat about.md

I'm a scientist-turned-engineer working at the intersection of synthetic biology, genomics, and agentic AI.

After a Ph.D. in systems biology at the University of Washington and five years at EMBL unpacking how genomes are organized and regulated, I moved into industry to build the computational tools that make biological design faster — first statistical and machine-learning models at Inscripta and Ginkgo Bioworks, and now AI agents at Generate:Biomedicines. I'm based in Albuquerque, NM, and work remotely with teams around the world.

tail -f now.log

Principal Scientist, Applied AI @ Generate:Biomedicines — Nov 2024–present

Shipping agentic products that reshape scientific workflows across the company:

cat experience.log

Nov 2024 – present

Principal Scientist, Applied AI

Generate:Biomedicines · Boston (Remote)

  • Shipped agentic products that reshape scientific workflows across the company
  • Architected agent SDK, MCP gateway, eval framework, CLI tooling, and a self-improving memory system
  • Built a digital-twin framework analyzing hundreds of thousands of DNA variants in parallel
Jan 2023 – Jun 2024

Associate Director, Genetic Design

Ginkgo Bioworks · Boston (Remote)

  • Founded a high-performing team and a new business function streamlining DNA asset management enterprise-wide
  • Prototyped AI-assisted knowledge management using RAG and embedding approaches
  • Predicted DNA synthesizability with gradient boosting, random forest, and logistic regression
  • Fine-tuned the HyenaDNA foundation model for regression and classification tasks
Dec 2020 – Dec 2022

Staff Scientist, Data Science

Inscripta · Boulder (Remote)

  • Enabled 4 customer publications and 3 internal patent applications
  • Led a team designing a biological knowledge graph for target selection
  • Developed statistical methods for interpreting multiplexed CRISPR selection experiments
  • Built an interactive Dash service for deep mutational scanning analysis and visualization
Jun 2015 – Nov 2020

Research Scientist

EMBL · Steinmetz Lab, Heidelberg

  • Secured €1.4M in funding; published 4 papers including Science and Cell, plus 1 patent
  • Led funding and regulatory approval for a clinical study using wearables to evaluate mitochondrial disorder symptoms outside the clinic
  • Built an end-to-end Snakemake pipeline processing terabytes of third-generation sequencing data
  • Applied gradient boosting to disentangle factors shaping transcript start/end sites in S. cerevisiae

ls ./projects

A handful of AI-native apps I've built for myself and my family, end to end — mostly local-first, each with a small embedded AI agent doing real work.

Also on GitHub — agentic dev tooling: agent-skills, agent-learn, agent-trade, pointy-rag →

cat publications.bib

13publications
852citations
12h-index

Full list on Google Scholar →

Patents

  • Methods for isolating and/or obtaining captured polynucleotide fragments US App. 18/501,530 · 2024
  • Genome-wide rationally-designed mutations leading to enhanced cellobiohydrolase I production in S. cerevisiae WO2023028521A1 · 2023
  • System and method for gene editing cassette design US20220246235A1 · 2022

Code behind the papers

  • scramble — analysis scripts for the Science paper on transcriptional neighborhoods

Education

  • Ph.D., Molecular and Cellular Biology — University of Washington 2008–2014
  • B.S. Biochemistry / B.A. Political Science — University of New Mexico 2002–2007

cat contact.txt

Always glad to talk agentic AI, genome engineering, or how one will unlock the other. The best way to reach me is a message on LinkedIn.