Paul R Saxman
AI and engineering leadership for global health and biomedical research
Seattle, WA, USA
Leading software and AI engineering for modeling frameworks at the Gates Foundation's Institute for Disease Modeling. My 20+ years of work experience across Google, Amazon, AWS, and academic medicine come together around a single arc: accelerating the science and decision-making that ultimately improves global health. M.Sc. Biomedical Informatics (UMich); co-author of 15 peer-reviewed papers with 4,600+ citations; 10+ years managing global cross-functional teams.
Open to advisory roles, speaking engagements, and collaborations in AI/ML, biomedical research, and global health.
saxman@gmail.com GitHub Hugging Face Google Scholar LinkedIn
Current Projects
- aimu A generative AI framework that I'm building to explore and understand how AI capabilities work together as a system. Supports multiple local and cloud model providers (Ollama, Hugging Face, Anthropic, OpenAI, Gemini) through a unified client interface with native extended-reasoning support. Includes autonomous agents, generation and understanding for text/language, images, audio, and voice, structured workflow patterns, MCP tool support, versioned prompt management and prompt tuning, memory systems, agentic skill support, and built-in evaluation and benchmarking.
- genscai Recipes and notebooks demonstrating how to apply generative AI to science and research workflows, with a focus on health and infectious disease modeling. Covers the full pipeline from automated literature retrieval (medRxiv, configurable APIs) through information extraction, document embedding and clustering with science-specific models (SPECTER, NV-Embed), document classification, knowledge graph construction, and intelligent agent development using agent frameworks and tools.
- hipai An exploration into how accurately a generative AI system can simulate a specific person, using ontological facts about the person and records of their style and presentation. Implements two applications: an AI assistant that retains personal context across conversations, and an AI clone that learns to respond in a user's voice from stored memories. Built on a vector memory store, a local MCP server, and pluggable LLM backends.
Experience
Roles
Management & Leadership
10+ years leading global and domestic teams across solutions architecture, software engineering, open source software, and developer relations.
- present Leading an AI and software engineering team at the Gates Foundation's Institute for Disease Modeling, delivering models, modeling frameworks, and AI systems that inform investment decisions for critcal disease and maternal/infant health programs.
- Led global technical programs at AWS for education and academic research customers worldwide, spanning K–12, higher education, and academic medicine.
- Led the US Academic Medicine Solutions Architecture team at AWS, composed of subject matter experts in EHRs, medical imaging, FHIR/OMOP, bioinformatics, HPC, and genomics.
- Built and managed cross-functional Developer Relations and Partnerships teams at Google for Android (wearables, TVs, tablets), Glass, Maps, Fit, and Health.
- Managed the engineering team at the University of Michigan Health System delivering software systems for clinical and translational research at MICHR.
Software Architecture & Engineering
Designing and delivering data-intensive, scientific, and clinical software systems with an emphasis on clean architecture, usable APIs, and user-centered design.
- present Architecting modeling and AI software frameworks for global health at the Gates Foundation, accelerating scientific discovery and policy decision-making.
- Led design and development of novel third-party data, AI, and ML services for Amazon Halo, Echo Buds, and Echo Loop (Amazon's first health and wellness consumer devices).
- Architected interoperable, HIPAA-compliant clinical research platforms and cross-institute federated research systems at the University of Michigan.
- Led architecture of Google's medical record interoperability services/APIs for Google Health, and developed full-stack, large-scale web applications, including Google's Flu Vaccine Finder.
- Designed repeatable, complex technical solutions leveraging a deep enderstadning of microservices, serverless, event-driven systems, full-stack web and mobile application architectures.
Solutions Architecture & Developer Relations
Helping enterprise customers and external developers build successful products on cloud and platform technologies, with a focus on health, science, and geospatial.
- Principal Solutions Architect at AWS guiding academic medicine customers on cloud architectures for clinical decision support, high-throughput genomics, and precision medicine.
- Managed strategic AWS initiatives around generative AI, machine learning, advanced computing, and federated research for academic institutions worldwide.
- Staff Developer Advocate for Google Maps, scaling developer messaging and supporting geospatial data visualizations and mobile mapping applications.
- Lead Developer Advocate for Google Health, supporting the platform's pivot to integrated clinical and device/wearable data, and represented Google at HHS ONC on PHRs in meaningful use.
- Managed launch-critical health & wellness developer partnerships for Google Wear, Glass, Maps, and Fit, and key Amazon Halo partners (Cerner, John Hancock, Weight Watchers).
Biomedical Research & Science
17+ years applying computing and data science to clinical research, epidemiology, microbiology, and global health.
- present Pioneering modeling and AI techniques at the Gates Foundation to enhance access and usability of infectious disease models for researchers and global health partners.
- Researched and developed data mining tools, visualizations, and ontology-based systems for clinical informatics and inter-institutional biomedical collaboration.
- Co-author of 15 peer-reviewed publications across biomedical informatics, microbiology, and clinical research, with 4,600+ citations.
- Designed and maintained widely used bioinformatics platforms during and after graduate study, including the Ribosomal Database Project, BSD, rrnDB, and T-RFLP analysis tools.
- Contributed to network analyses of cancer symptoms and renal disease genes, advancing molecular-based disease classification.
Full work history
Driving improvements in global health by delivering novel infectious disease and health models and software frameworks that directly inform the Gates Foundation's strategy and investment decisions for malaria, HIV, polio, maternal/infant health, and other priority programs. Pioneering modeling and AI techniques that enhance model access and usability for researchers and global health partners.
Led global technical programs to accelerate adoption of cloud services by education and academic research institutions worldwide. Developed strategic initiatives around generative AI, machine learning, advanced computing, and federated research for K–12, higher education, and academic research customers.
Led Academic Medicine Solutions Architecture for the US market. Team of subject matter experts in EHRs, medical imaging, healthcare data standards (HL7, FHIR, OMOP), bioinformatics, HPC, genomics, and AI/ML. Guided customers on optimal cloud architectures and complex healthcare and biomedical research problems including clinical decision support and precision medicine.
Led design and development of novel third-party data, AI, and ML services for Amazon Halo, Echo Buds, and Echo Loop (Amazon's first health and wellness consumer devices). Managed technical partnerships with Cerner, John Hancock, and Weight Watchers.
Led global, cross-functional Android Devices, Media & Fitness Developer Relations team responsible for Android device launches and support, including Google Wear, Google Fit, and Google TV.
Managed global Health & Wellness Developer Partnerships team responsible for launch-critical partnerships for Google Wear, Glass, Maps, and Fit. Led third-party go-to-market strategy for all Google product launches with health & wellness use cases.
Focused on mobile app development, geospatial data visualizations, and helping developers create compelling map applications at scale.
Led technical integrations as Google Health pivoted from a personal health record to a health & wellness platform. Represented Google at HHS ONC on PHRs in meaningful use. Led architecture of medical record interoperability APIs and developed Google Flu Vaccine Finder.
Advanced clinical and translational sciences through the design and development of interoperable, HIPAA-compliant systems at the Michigan Institute for Clinical & Health Research (MICHR). Architected cross-institute federated research platforms and managed the engineering team delivering software systems for clinical and translational research.
Featured Work
- Whitepaper Modernizing your healthcare-data strategy AWS Prescriptive Guidance, November 2023. Procedural, organizational, and technical guidance for healthcare executives advancing a data-driven mission, covering interoperability, security and compliance, research enablement, and cost sustainability, framed around the Quadruple Aim.
- Blog Post HMS People Heart Study streamlines digital research with AWS AWS Industries Blog, June 2025. How Harvard Medical School's People Heart Study uses AWS HealthLake and connected services to ingest patients' provider records and deliver personalized cardiovascular-risk insights through an iPhone app.
- Blog Post Generative AI in education: Building AI solutions using course lecture content AWS Public Sector Blog, October 2023. Reference patterns and architecture for turning recorded course lectures into generative AI-powered learning experiences, including summarization, homework generation, chatbots, and personalized learning features for higher education.
- Blog Post Germany's International University of Applied Sciences automates creation of educational videos using generative AI, serverless on AWS AWS Public Sector Blog, May 2024. Customer case study on how IU, Germany's largest university with 130,000+ students, replaced a manual video production workflow with a serverless, generative-AI pipeline capable of producing the 24,000 educational videos needed for 2024.
More featured work (10)
- Blog Post Hummingbird: a tool for effective prediction of performance and costs of genomics workloads on AWS AWS Industries Blog, March 2022. Introduces Hummingbird, an open-source Stanford tool that profiles genomics workloads and recommends the fastest and most cost-effective EC2 instance types, illustrated with BWA-MEM and GATK HaplotypeCaller workflows.
- Blog Post Improving the utilization of wearable device data using an AWS data lake AWS Industries Blog, February 2021. How healthcare organizations can build an AWS data lake to ingest, normalize, and analyze wearable-device data across manufacturers and sampling rates, enabling research aggregation and machine learning on patient-generated health data.
- Webinar Texas A&M University's approach to modernizing research through the cloud AWS IMAGINE for Education, January 2021. Fireside chat with Dr. Joshua Kissee on STAR (Secure Technologies for Aggie Researchers), Texas A&M's self-service cloud platform for research computing, secure data collaboration, and AI/ML workloads.
- Lecture Research Foundations on Amazon Web Services Stanford Deep Data Research Center, July 2022. Academic lecture introducing the foundations of running research workloads on AWS, oriented toward academic research teams and PIs new to cloud-based scientific computing.
- Conference Talk Building Rich Media Experiences with Android Google Developer Days India, December 2017. Co-presented with Dan Galpin on building outstanding media apps for Android, covering the Android Media Framework, picture-in-picture, MediaStyle notifications, the Channels and Programs API for Android TV, and high-performance audio for media creation apps.
- Webinar UXD: What is User Context and Why Does it Matter? Android Developers, January 2014. Episode in the User Experience Design for Developers series, distilling user-research data on how people use their devices, the expectations they bring, and how those expectations shift with context.
- Webinar Google Maps API: Snazzy Maps Google Developers, December 2013. Spotlight on Snazzy Maps, a community-driven gallery of styled-map presets for the Google Maps API.
- Conference Talk Making Location Meaningful with the Google Maps APIs Google I/O 2013. Co-presented with Kasia Derc-Fenske on going beyond raw position by using destinations, history, and behavior to deliver richer location-aware experiences with the Google Maps APIs.
- Conference Talk Command and Control in the Living Room: Building Second Screen Apps for Google TV Google I/O 2012. Co-presented with Dave Fisher on the second-screen pattern for Google TV: using smartphones, tablets, and PCs as companion controllers and content sources for the living-room device.
- Conference Talk Google I/O 101: Introduction to Google TV Google I/O 2012. Introductory session walking developers through the basics of building for Google TV, the Android-powered device at the center of the living room.
Research
Projects & Systems
- Biomedical Resource Ontology (BRO) & Resource Discovery System Federated, inter-institutional ontology and discovery system enabling semantic annotation and search of biomedical resources across CTSA institutions.
- University of Michigan Honest Broker Web Services-based platform for sharing de-identified clinical and research data across medical disciplines and institutions while preserving patient privacy.
- VISAGE Visual Aggregator and Explorer: a query interface for clinical research with ontology-driven query building, management, and exploratory data mining for cohort identification.
- Ribosomal Database Project (RDP-II) Aligned and annotated rRNA sequence database with phylogenetic tree derivation and analysis services, supporting microbial taxonomy worldwide.
- Biodegradative Strain Database (BSD) Web-accessible database of degradative bacteria and the hazardous substances they degrade, with linked literature, patents, and chemical data.
- Ribosomal RNA Operon Copy Number Database (rrnDB) Phylogenetically arranged database of rRNA operon copy number across prokaryotes, supporting evolutionary and microbial community analysis.
- Terminal Restriction Fragment Length Polymorphism Analysis Program (TAP T-RFLP) Web-based research tool for in silico restriction digestion of 16S rDNA sequences, enabling microbial community analysis and genetic fingerprinting.
Publications (15)
-
Bahmani A, Cha K, Alavi A, et al. "Achieving inclusive healthcare through integrating education and research with AI and personalized curricula"
-
Tenenbaum JD, Whetzel PL, Anderson K, et al. "The Biomedical Resource Ontology (BRO) to enable resource discovery in clinical and translational research"
-
Zhang GQ, Siegler T, Saxman P, et al. "VISAGE: a query interface for clinical research"
-
Bhavnani SK, Bellala G, Ganesan A, et al. "The nested structure of cancer symptoms"
-
Bhavnani SK, Ganesan A, Hall T, et al. "Discovering hidden relationships between renal diseases and regulated genes through 3D network visualizations"
-
Mirel BR, Wright Z, Tenenbaum JD, Saxman P, Smith KA. "User requirements for exploring a resource inventory for clinical research"
-
Boyd AD, Saxman PR, Hunscher DA, et al. "The University of Michigan Honest Broker: a Web-based service for clinical and translational research and practice"
-
Bhavnani SK, Eichinger F, Martini S, Saxman P, et al. "Network analysis of genes regulated in renal diseases: implications for a molecular-based classification"
-
Patel R, Hanratty M, Johnson J, Saxman P, et al. "Casepedia: a Web 2.0 case repository enabling collaborative learning"
-
Boyd AD, Hunscher DA, Kramer AJ, et al. "The 'Honest Broker' method of integrating interdisciplinary research data"
-
Urbance JW, Cole J, Saxman P, Tiedje JM. "BSD: the biodegradative strain database"
-
Klappenbach JA, Saxman PR, Cole JR, Schmidt TM. "rrndb: the ribosomal RNA operon copy number database"
-
Maidak BL, Cole JR, Lilburn TG, et al. "The RDP-II (ribosomal database project)"
-
Maidak BL, Cole JR, Lilburn TG, Parker CT, Saxman PR, et al. "The RDP (ribosomal database project) continues"
-
Marsh TL, Saxman P, Cole J, Tiedje J. "Terminal restriction fragment length polymorphism analysis program, a web-based research tool for microbial community analysis"
Credentials
Education
-
M.Sc. Biomedical Informatics, University of Michigan Medical School, 2006–2008
Clinical informatics, complex systems, and network theory -
B.Sc. Computer Science & B.Sc. Computer Engineering, Michigan State University, 1994–1999
Cognate in Biomedical Engineering. Co-authored research papers with 4,600+ citations.
Certifications
- AWS Certified Solutions Architect, Professional
- AWS Certified Machine Learning, Specialty
- AWS Certified Developer, Associate
- AWS Certified Cloud Practitioner
Patents
-
"Systems and Methods of Using a Temporary Private Key Between Two Devices"
2012