Jay Fowler
Finance, grants, and AI systems for organizations that run on public money and need to run better.
Co-founder & CFO, Sky Island AI, Tucson, AZ
Selected Impact
$20M+
Annual recurring federal grants won
~$90M
Total projected value across multi-year awards
$70M
Budget managed with FAR/GAAP compliance
30+
Media mentions for Project Blue impact dashboard
Services
Fractional CFO & Finance
I've run budgets up to $70M with FAR and GAAP compliance, built KPI reporting for boards, and tightened internal controls at organizations where the finance function needed senior leadership but not a full-time hire.
AI Systems & Automation
I build multi-agent systems, real-time voice interfaces, and workflow automation that actually go to production. Sky Island AI just won a City of Tucson grant to deploy our Virtual Case Manager for tenant support.
Grant Strategy
I've written and won $20M+ in annual recurring federal grants. I know the full cycle: finding the opportunity, writing the proposal, managing the budget, filing the reports, and renewing the award.
How I work: Retainer or project-based. I embed with your team, build what's needed, and hand off clean. Contracts go through Sky Island AI.
Selected Work
Virtual Case Manager
AI case management platform with persistent memory across sessions. Handles Medicaid, courts, and social services. Won a City of Tucson grant to deploy it for tenant support.
Python, Real-time Voice, Multi-agent
Economic impact dashboard for Tucson's proposed data center. Cited in national news coverage. 30+ media mentions and 10k+ views.
Data Modeling, Public Policy
Workflows
Multi-agent workflow orchestration across Claude Code and Codex CLI. Tri-model convergence framework for high-stakes decisions.
Python, Multi-LLM Orchestration
single-shot-code
public9-step code generation pipeline with tri-model convergence, scope enforcement, and automated testing.
Shell, Multi-model Pipeline
medical-diagnostic
publicIterative tri-model diagnostic system with anti-anchoring and human-in-the-loop validation.
Multi-model, Safety-critical
Prompt Testing Framework
publicBlind pairwise A/B testing framework for prompts across model families. Negative controls, ablation studies, cross-model evaluation.
Prompt Engineering, Evaluation
proofer-extension
publicLocal-first AI proofreading for Chrome and Brave. Runs on Ollama, no data leaves your machine.
JavaScript, Ollama, Chrome Extension
Interactive data analysis of Arizona's 2024-25 fencing season. 1,596 competition instances, 398 fencers, 133 events.
Data Visualization, GitHub Pages
BladeStack
Fencing tournament management system. Offline-first architecture with real-time sync and integrated refereeing.
TypeScript, Offline-first
About
I got my degree at the University of Arizona, double major in economics and business administration, plus an honors thesis in chemistry that I still can't fully explain the career relevance of. After a couple years running materials and operations at a medical device company, I landed at VisionQuest National, which was a 100% grant-funded for-profit. Unusual setup. I was there seven years, eventually managing the full $70M national budget across federal, state, and local sources. Reported to the CFO and CEO, worked directly with the Board chair. Got pulled into a four-month stint running the development office, wrote about 600 pages of proposals, and brought in over $20M in annual recurring grants.
I've chaired the Arizona Division of USA Fencing since 2018. It's a USOPC member organization, around 500 members. I also chair the Still Point Fencing Foundation, which I set up as a 501(c)(3), and I coach there, running a staff of 11. My first language growing up was ASL. I have three deaf and hard-of-hearing sisters, which is part of why accessibility and communication matter to me beyond just checking boxes. I've traveled to 39 countries at this point, mostly because I'm restless.
Contact
If any of this sounds like it could help your organization, reach out.
Contracts go through Sky Island AI.