For recruiters & hiring teams

Ryan Booth

Applied AI Architect Canyon, TX · Remote-first boothr2@gmail.com

Architecture and leadership in the trenches — building applied-AI products and agentic systems, and helping teams ship solid solutions with AI tooling, not AI theater.

Tailor this to your role → Download PDF LinkedIn

Hiring me as a consultant instead? That's /consult — different page, different engagement.

Answers before the screening call

A recruiter screen is mostly logistics — fit, location, authorization, timing, money. There's no reason you should have to spend a call collecting them. Here they all are. If something below is a dealbreaker, we both just saved a week.

Target roles
Applied AI Architect · Applied AI Specialist · AI Solutions Architect · AI Engineering Lead / Manager
Role shape
Senior IC or hands-on lead. I want to be in the code and the architecture, not managing away from it. I have run a team of five and would do it again — as a player-coach, not a pure people manager.
Location & travel
Canyon, Texas (US Central). Remote-first, and I have worked remote-first since 2017. Happy to travel for onsites, planning weeks, and customer work. Not looking to relocate.
Work authorization
US citizen. No sponsorship required, now or in the future.
Availability
Available now. Independent through Blue Ridge Systems, so there is no notice period to work through — I can start on a normal onboarding timeline.
Compensation
Tell me the band and I'll tell you straight if it works. I would rather find that out on the first call than at offer stage, and I am not going to waste your pipeline time.
What I want
Real problems with real constraints, and a team that ships. I have spent the last two years going deep on agentic systems in production, and I want to do that work somewhere the results matter to a business — not in a lab, and not as AI theater.

Resume tailoring

Point it at your job description

Paste the description and you'll get a match report — which of my actual bullets are evidence for which of your requirements, and where I don't line up — plus a resume ordered for this specific role.

It reorders and selects from resume content I wrote and can defend. It never rewrites a bullet or invents a metric, so nothing here will contradict me in an interview. Job description text isn't stored.

or

Best effort — LinkedIn, Indeed and Workday block automated reads. Pasting the text always works.

Or just download the resume

Selected Projects

2025 – Present

Tokru.ai — Founder / Architect

Independent R&D into agent-native professional and brand presence. Live today: a capability registry that turns a person's existing work into a structured, machine-readable profile — AI-drafted, published at {handle}.tokru.ai, and interactive by both humans and AI agents over MCP and A2A. Roadmap extends the same own-your-presence model to makers, creators, and artists, with A2A/A2P powering agent-driven purchasing.

tokru.ai →

2025 – Present

Root Cause — agentic-development case study

A datacenter-simulation game — power, thermal, CLOS networking, cascading failures, a local multi-agent AI narrator — built in Rust + Bevy, a stack I had never used, with 80+ integration tests shipped. AI executed under my architectural direction. Released free / pay-what-you-want on itch.io. The point was the methodology; the game is real.

abstractryan.itch.io/root-cause →

2024 – 2025

Artist Dashboard

Full Python/FastAPI SaaS built entirely via AI prompts with no hand-written code. The product shut down; the methodology became the foundation for Root Cause.

2019 – 2021

apstra-api-python

Open-source client library for the Apstra REST API, written to move network engineers from CLI workflows into API-driven automation — and to teach while it did.

Work Experience

Nov 2024 – Present

Founder & AI Solutions Architect — Blue Ridge Systems, LLC

Independent AI architecture and applied R&D. The LLC is the umbrella for consulting and product work; the substance lives in the projects.

  • Design and build agentic workflows, orchestration patterns, and AI-native applications end to end — across MCP, agent-to-agent (A2A), and agent-to-payment (A2P) protocols.
  • Developed a repeatable agentic-development methodology: building production-grade systems with AI executing under architectural direction — proven on a full SaaS product and on a complex simulation written in an unfamiliar language.
  • Built and operate a personal multi-agent harness (Lisa) and a self-hosted observability stack (Milhouse) covering all development and applications, fully operable through AI.
  • Run local and secure model serving (Ollama, vLLM) alongside hosted models; build custom MCP servers and agent tooling.

Jan 2021 – Jun 2025

Engineering Manager — Juniper / Apstra

Owned and scaled CloudLabs through the Juniper acquisition; in later years also drove internal AI pathfinding.

  • Led a team of ~5 engineers building and operating CloudLabs, a globally distributed lab and demo platform used across sales, support, product, and engineering.
  • Survived the Juniper acquisition and made CloudLabs the platform that delivered CEO Rami Rahim's mandate to ramp the entire sales and support organization on Apstra within the first quarter post-close — driving a 500% increase in adoption post-acquisition and ~10% year-over-year growth against a baseline agreed with Apstra leadership.
  • Defended the platform repeatedly against better-resourced internal alternatives — winning on merit with a four-to-five-person team.
  • Cut environment provisioning from the enterprise norm of weeks or months to hours — five dedicated lab environments stood up in about two hours — by integrating CloudLabs into engineering's Jenkins CI/CD pipeline, putting a fresh product build in front of a customer within hours of release.
  • Dedicated roughly 20–40% of later-year capacity to internal AI pathfinding — partnering with engineering teams to prototype and evaluate agentic and RAG-style workflows; deepened this independently after hours, which became my full-time focus on leaving Juniper.

Jul 2017 – Jan 2021

Sr. Automation Solutions Architect — Apstra

Established the platform foundation across software, infrastructure, virtualization, and user provisioning before the Juniper acquisition.

  • Took CloudLabs from a handful of internal Ansible scripts to a real, customer-facing cloud platform; scaled the team from 1 to ~5 engineers and rebuilt the architecture for growth.
  • Architected the platform: Flask + Celery + Redis backend, React + TypeScript frontend, and KVM-based virtualization of multi-vendor network hardware (Cumulus, Arista, Cisco, Juniper) across multiple AWS regions on bare metal.
  • Auto-provisioned external users via HubSpot CRM and internal users via Google Workspace SSO — making the sales motion and product access a single motion.
  • Built and open-sourced apstra-api-python, a client library that met network engineers where they were (CLI → REST) and taught while it did.
  • Bridged networking and software-engineering teams to enable customer automation strategies.

Nov 2015 – Jul 2017

Sr. Network Engineer — PNC Bank

Where network automation became the career.

  • Operated and helped build the core for a nationwide bank: five interconnected datacenters on a 100GbE MPLS backbone (Cisco ASR routers, Nexus datacenter pods) carrying WAN, internet, and all US banking operations.
  • Stood up a network automation function: built data, alerting, and troubleshooting tooling so the datacenter engineering team could reach root cause faster; automated config builds, change execution, and data collection with Ansible.
  • Led a year-long evaluation to migrate the MPLS DCI (IOS-XR) toward Segment Routing and BGP-LS, targeting more resilient, programmatic failure routing. Reached UAT; not productionized before departure.
  • Early adopter of streaming telemetry — experimental Arista CloudVision work when the technology was brand new — which led directly to being recruited into Apstra.
  • Evaluated automation vendors and startups and drove the selection of Network to Code for the datacenter engineering organization.

Sep 2013 – Nov 2015

Sr. Network Engineer — BSA Health System

Origin story; where the CCIE was earned.

  • Joined the team remediating a hospital network degraded by years of technical debt and failed rebuild attempts — which had culminated in a 3–4 day system-wide outage severe enough to force patient diversions.
  • Partnered with consulting firm Bedrock to root-cause the failure — a misarchitected core producing spanning-tree loops and routing instability — then designed and executed the full rebuild.
  • Built a proper datacenter fabric (Cisco Nexus 7K/5K/2K), re-established a stable enterprise campus core (6500 VSS + 3850 access), and stabilized every branch and secondary hospital site.
  • Standardized switch, router, and firewall templates; integrated multiple clinic acquisitions with a new remote-site model; mentored teammates.
  • Earned CCIE Routing & Switching #50117 during this tenure (2015).

Jun 2009 – Sep 2013

Security Specialist — B&W Pantex

  • Designed and deployed a secure, redundant core network in a high-availability, mission-critical environment (Cisco Nexus + Palo Alto firewalls); device-hardening standards; Motorola outdoor wireless mesh across a 16,000-acre plant; RADIUS authentication and monitoring.

Check My Work

Screening is a trust problem, so here are the primary sources. Nothing on this page depends on you taking my word for it.

Tech Field Day →

Delegate at NFD8 (2014) and NFD12 (2016); solo presenter at NFD19 (2018) and NFD28 (2022); delegate and presenter at AI Infrastructure Field Day 8 on A2A/A2P. Delegates are community-selected independent practitioners — the selection is the credential.

Podcasts

Kentik Telemetry Now (4 episodes — vibe coding, LLMs in network ops, demystifying GenAI/LLMs); Packet Pushers Ep. 311; Searching for Null0 (building apps with AI); The CTO Advisor panel, "Ready or Not, AI is Coming to the Enterprise."

AIMUG Austin →

Ongoing contributor and presenter — agentic systems, RAG, and LLM tooling.

Publishing →

abstractryan.ai; a paid Substack on AI and infrastructure (including a two-part AI model-router series); YouTube devlog and GitHub as @that1guy15.

Code

github.com/that1guy15 — including Milhouse (Apache-2.0, active) and apstra-api-python. This site is its own case study, and it answers to agents at /a2a.

Still reading?

Then it's probably worth a conversation. Bring the job description — or run it through the tailor above first and we can start from what actually lines up and what doesn't.

Get in touch Book a call