Anthony Foster

Staff Software Engineer & Architect | Distributed Systems | Agentic AI Strategy & Governance

Detroit Metropolitan Area amfoster.dev@gmail.com LinkedIn

Summary

I've spent the last 15+ years building distributed systems that have to work. That includes large-scale data platforms moving billions of events, and now software that supports clinical trials for life-saving treatments.

Lately, I've been spending a lot of time thinking about what it takes to operate AI agents safely in production, especially as these systems become longer-lived and more embedded in real workflows.

I'm currently a Staff Engineer and Tech Lead on a high-compliance, AI-powered EMR-to-EDC platform. I work closely with my team on the design and delivery of complex clinical data workflows in a strict regulatory environment (CFR Part 11), with an emphasis on enabling agent-assisted development workflows that make sense for regulated systems.

I'm interested in autonomous systems with a focus on governance, observability, and auditability over time. I write about AI, architecture, trust, and governance at tonyfoster.dev.

Skills

TypeScript/Node.js C#/.NET Kotlin/Java React Next.js AWS GCP Kafka Kubernetes Terraform OpenTelemetry Grafana PostgreSQL MongoDB Docker CI/CD

Experience

Staff Software Engineer

Oct 2025 – Present

Paradigm Health · Columbus, Ohio

  • Partnering with Engineering Leadership and Product to define the AI DevEx roadmap; establishing standards for internal AI tooling and development practices.
  • Serving as Tech Lead for the AI-powered Data Extraction and Automated EHR-to-EDC platform; driving the architectural evolution of a high-compliance, multi-service ecosystem to support complex clinical workflows.
  • Architected and implemented the platform's immutable audit logging system to support CFR Part 11 compliance; establishing a verifiable data tracking framework to ensure regulatory trust for critical clinical assets.

Software Engineer

Mar 2025 – Sep 2025

Curipod · Oslo, Norway

  • Built and maintained Curipod's interactive, AI-native curriculum & student-participation platform, Next.js frontend with a TypeScript/Koa backend over HTTP and WebSockets on GCP, scaling to hundreds of thousands of users daily.
  • Lifted signup conversion to 96% (from 90%) by redesigning onboarding and running analytics-driven experiments.
  • Established a culture of observability with OpenTelemetry, instrumenting HTTP and WebSocket services to monitor AI-agent and LLM performance; shipped self-serve debug dashboards accelerating feedback loops.
  • Automated release processes, removing manual steps and saving ~7 engineering weeks per year.
  • Ran in-person and remote user tests with teachers and students; partnered with product/design to turn insights into specs and shipped features more quickly.

Principal Software Engineer, Platforms Team Lead

Aug 2021 – Mar 2025

SkySpecs · Ann Arbor, Michigan

  • Improved platform observability, resulting in a 2–3x decrease in debug time, by creating TypeScript instrumentation libraries and using Terraform to provision supporting infrastructure. Processing 500+ GB of log, metrics, and tracing data monthly.
  • Built and managed the data streaming platform leveraging PostgreSQL, Kafka Connect, and RedPanda, deployed on Kubernetes with Helm. Streams billions of events across services and into a data warehouse.
  • Designed and built a notifications system on the data streaming platform in collaboration with product management, leveraging an event-driven architecture to alert customers of critical wind turbine damage.
  • Audited AWS infrastructure and source code for security vulnerabilities in collaboration with the information security team, contributing to the attainment of ISO 27001 certification.

Software Engineering Consultant

Oct 2020 – Jan 2024

Ripple Science · United States

  • Enabled scaling of HIPAA-compliant clinical research participant management software for enterprise users, helping secure contracts with record-setting ARR by migrating authentication to Amazon Cognito.
  • Achieved a 98% improvement in MongoDB query times by refactoring the schema and rewriting aggregation pipelines, resolving critical performance bottlenecks.
  • Reduced deployment-related issues by 90% by introducing canary deployments.
  • Developed a participant communication framework leveraging SendGrid and Twilio APIs, helping achieve a 96% clinical trial retention rate (far exceeding the ~70% industry standard).

Senior Lead Software Engineer

Feb 2021 – Aug 2021

Coupa Supply Chain, powered by LLamasoft · Ann Arbor, Michigan

  • Reduced supply chain model development time from 3–6 months to minutes by designing a system for automatic model generation.
  • Increased deployment speed and confidence by developing an integration test framework in C# using .NET Core, leading to widespread adoption across the platform.

Software Architect

Oct 2018 – Oct 2020

Coupa Supply Chain, powered by LLamasoft · Ann Arbor, Michigan

  • Architected data ingestion APIs using C# (.NET Core) on Kubernetes with Apigee API gateway. Built a data ingestion pipeline on Apache Kafka, enabling onboarding of Fortune 500 clients.
  • Created incident management protocols by leading adoption of processes inspired by the Atlassian Incident Management handbook. Mentored engineering leaders on proper root cause analysis.

Senior Software Engineer, ESP

Apr 2017 – Oct 2018

Coupa Supply Chain, powered by LLamasoft · Ann Arbor, Michigan

  • Led full-stack engineering for the Enterprise Simulation Platform, featuring a React frontend and C# (.NET Core) backend on Kubernetes with embedded Tableau Server dashboards.
  • Mentored cross-functional interns through delivery of a prototype for embedded analytics software, guiding requirements gathering, agile practices, and deployment automation.

Education

Michigan State University

Bachelor of Science — Computer Science, Computational Math