Andrew Weston
Open to work
Forward-deployed / solutions engineer  ·  Applied AI

I deploy AI inside real companies, and I help build lasting systems that keep it working.

Fourteen years of hands-on work, from a computer repair bench to deploying clinical software inside hospitals. I pride myself on careful and methodical integration work. The software suite I deploy today includes multiple AI systems. The systems I build and run on my own time do too. I enjoy knowing both sides, deploying AI-first workflows for others and building them for myself.

01  Experience

2020 —
Installation Systems EngineerLunit

Deploy a computer-vision AI radiology software suite across hospitals and imaging centers, and own integration and functionality end to end. Design and implement software integrations; primary technical liaison between system administrators, analysts, and developers.

2017 – 20
IT Support EngineerVolpara Solutions

Ran HIPAA-regulated PHI environments on MS SQL. MS365 global admin (Azure AD, Exchange, Teams) for a hybrid cloud; managed on-prem servers and Azure VMs with PowerShell and automated recurring processes.

2015 – 17
Ultrasound Systems Repair ManagerSummit Imaging

Led a team of six technicians; wrote the work instructions and shop procedures that earned ISO 9001 and 13485 certification, the same documentation discipline agentic AI governance demands. National travel to resolve complex system failures.

2012 – 15
Repair TechnicianSummit Imaging / PC Fix

Ultrasound component repair and retail computer hardware/software diagnosis, direct customer support. Where the habit of fixing things at the client's site started.

02  Cornerstones

Two of the systems I build and operate outside the day job.
Platform
DeviceClearance.com

252,600 pages of FDA device-clearance and recall data, rebuilt on Cloudflare Workers + D1 after hitting Pages' static-file limit.

Live since April 2026
OS
System O

A markdown-based OS for AI-augmented operators: vault scaffolding, offline automation chain, plugins, and dashboards.

In progress

03  Background

I started out fixing hardware and software problems in retail and worked up to deploying clinical software inside hospital systems: HL7 interfaces for reporting software, integrations with dictation systems and PACS, plus the HIPAA governance and paper trails regulated software demands. Deploying an AI suite in that world taught me what it takes to make software behave when real people depend on it.

Now I build my own systems from scratch: LLM agents, MCP servers, and the operational plumbing that keeps them running. I've seen how AI gets deployed and how it gets made, and I'm looking for forward-deployed or solutions roles where doing both matters.

TypeScript / Python / Rust MS SQL Server · HL7 HIPAA / PHI governance CompTIA Security+ Active DoD Confidential clearance Agent systems · MCP Claude Code · Codex CLI · OpenCode Azure · Google Cloud · Hybrid Local model deployment

04  Security posture

Small attack surface, explicit controls, verifiable claims.

This site ships no framework, package dependencies, trackers, cookies, forms, or third-party runtime code. A default-deny Content Security Policy and defense-in-depth response headers constrain what the browser may execute, load, frame, or disclose.

Security posture & disclosure →

Open to the right role.

Early-stage applied-AI teams especially. If you need someone who can deploy the product and build it, send me a note.