John Parker, PhD

Leading software engineering, data platform, and AI strategy at a biotech startup, aligning cloud infrastructure and AI initiatives with business goals

• Established Agentic AI systems on AWS Bedrock and Databricks using Claude Agent SDK, Agent Skills, and MCP, with knowledge base connectors to Confluence and electronic lab notebooks
• Drove company-wide Claude Enterprise adoption as a change management program across Data Science, Lab Departments, IT, and Business, anchored by security guardrails, responsible-use training, and PR and documentation governance gates
• Managed cross-platform budget forecasting and scaling across AWS, Databricks, and Claude AI, including spend attribution by department
• Established company-wide data strategy and SOPs consolidating fragmented lab and sequence data into a unified bronze/silver/gold pipeline on S3, Databricks Declarative Pipelines, and Lakebase

Managed a software engineering team at a biotech startup, overseeing cloud-based deep learning and bioinformatics workflows while providing platform support across data scientists, lab scientists, and leadership

• Led company-wide migration to the Databricks platform, centralizing data and enabling self-service analytics
• Partnered with IT to enable real-time lab data ingestion of many instruments using AWS S3 File Gateway and Spark Declarative Pipelines
• Transitioned team focus from infrastructure maintenance toward application and data engineering with direct business impact

Developed and maintained a greenfield machine learning software platform on AWS to enable and support Data Science workloads

• Architected greenfield ML platform on Kubernetes (EKS) with Metaflow and Argo Workflows, supporting parallel GPU training jobs for VAE and transformer models
• Unified ML/SWE repositories into a trunk-based monorepo with GitHub Actions CI/CD, merge queues, and feature flags for continuous deployments
• Migrated data warehouse to Apache Iceberg lakehouse with FastAPI/Panel stack for dataset streaming and cross-team data sharing
• Partnered with NVIDIA as early access collaborator to integrate BioNeMo T5 LLMs into custom VAE architecture on Kubernetes
• Contributed to Netflix's open-source Metaflow, resolving an issue in its Argo Workflow generator

Provided software development, training, and system administration for a Linux-based high performance computing (HPC) cluster serving University researchers

• Benchmarked CPU and GPU performance on next-gen HPC cluster with NVIDIA A100 (MIG), AMD, and Intel hardware using CUDA and MPI.
• Built multi-GPU molecular dynamics framework using JAX and MPI, scaling ensemble simulations beyond single-node limits.
• Led interactive HPC workshops on SLURM, parallel Python/MPI, and high-performance computing concepts.

Conducted computational physics research spanning simulation development, software publication, and academic scholarship.

• Published cross-platform Python & C++ simulation packages for electrodynamics and hydrodynamics, using CMake and pybind11
• Co-authored over a dozen peer-reviewed publications with collaborators at the University of Chicago, Argonne National Laboratory, and partner academic institutions.

PhD Physics

Computational physics research in Norbert Scherer's Lab. Thesis "Collective Electrodynamic Excitations and Non-conservative Dynamics in Optical Matter and Meta-atom Systems"

BS Physics, Minor in Mathematics

Computational physics research in Daniel Goldman's Complex Rheology And Biomechanics (CRAB) lab, studying the physical properties of granular materials.