CV

Current CV for Noah Flynn, Senior Research Scientist at Google Cloud AI.

Experience

  • Apr 2026 – Present
    Senior Research Scientist
    Google Cloud AI, Gemini Enterprise, Santa Clara, CA
    • Building agentic AI systems for Gemini Enterprise research, coding, and data science workflows.
    • Developing evaluation, adaptation, and reliability methods for production AI assistants that reason over long context and use tools.
  • Sep 2024 – Apr 2026
    Applied Scientist II
    Amazon AWS AI Labs, Agentic AI, Santa Clara, CA
    • Built customizable multi-agent primitives that shortened internal agent-development cycles for deep research, data science, database analysis, code transformation, and security-analysis workflows.
    • Co-developed agent-as-a-judge evaluation methods for personalized assessment of deep research agents with internal AWS teams and enterprise customers.
    • Engineered cost-aware reasoning and tool-use optimization for agents working with licensed, premium data sources.
    • Mentored 5 applied scientists on agent systems, evaluation methodology, dynamic context optimization, and parameter-efficient reinforcement learning.
  • Jun 2024 – Present
    Adjunct Faculty
    UC Berkeley, Berkeley, CA
    • Teach CHEM 274B, a graduate course on software engineering, machine learning, and cheminformatics for molecular science.
    • Mentor UC Berkeley MSSE capstone teams on applied machine learning and scientific software projects.
  • Jun 2021 – Sep 2024
    Research Scientist II
    Amazon Alexa & AGI, Santa Clara, CA
    • Accelerated Alexa's expansion into 3 international markets by developing dynamic data selection methods that reduced multilingual NLU data requirements by 95%.
    • Shipped and maintained 20 production releases of conversational AI models serving 100M+ monthly active users across 6 international markets.
    • Contributed tool-use capabilities and API invocation strategies to the Amazon Nova foundation model family and Alexa+.
    • Built calibration methods that improved robustness to uncertainty from upstream ASR transcription artifacts.
  • May 2019 – Aug 2019
    Machine Learning Intern
    Merck, Kenilworth, NJ
    • Built generative models for de novo drug design integrated with docking pipelines.
    • Explored reinforcement learning for targeted optimization of binding affinity and therapeutic property profiles.
  • Aug 2017 – Jun 2021
    PhD Candidate
    Washington University in St. Louis, St. Louis, MO
    • Modeled drug metabolism and bioactivation networks with deep learning methods, including graph neural networks, to guide early-stage drug development.
    • Developed and maintained XenoSite and XenoNet prediction web services for small molecule biochemistry.
    • Developed signal detection methods for retrospective electronic health record analysis of drug-drug interactions and drug-induced liver injury.
  • May 2016 – May 2017
    Computational Research Fellow
    NCSA, University of Illinois at Urbana-Champaign
    • Characterized proteomic and transcriptomic networks with graph-theoretic methods for engineered biological systems.
  • Nov 2015 – May 2017
    Software Engineer Intern
    AbbVie, North Chicago, IL
    • Developed a bioinformatics visualization suite for drug-target interactions, protein-protein interactions, and regulatory networks.

Education

  • Aug 2017 – May 2021
    PhD, Deep Learning & Computational Biology
    Washington University in St. Louis, St. Louis, MO
    • GPA 3.94/4.00
    • Dissertation on graph neural networks for drug metabolism prediction.
  • Aug 2013 – Jun 2017
    BS, Bioengineering — Minor in Computer Science
    University of Illinois at Urbana-Champaign, Champaign, IL
    • GPA 3.70/4.00