AI can build almost anything now. Knowing what to build, what good looks like, and what people will actually use is the work. When execution is no longer scarce, judgment is.
Work
Real estate and property operations. Systems people actually use.
Leading AI strategy at Connexus, a multifamily procurement and vendor management platform. Architecture, agent tooling, and the data workflows behind procurement intelligence.
AI-powered company research and intelligent roadmap for Shadowpoint Media. Driving data-backed marketing decisions across their real estate media and creative services portfolio.
CFG Reserve, the Tucson hospitality and short-term rental company I founded after Panoply. In-house operations, pricing that moves hourly, and an AI-native back office that lets a small team run like a large one. 4.97 average guest rating.
Where the operator instincts come from. Co-founded Panoply and scaled it to 45+ short-term rental properties at over 90% occupancy before a strategic exit. Pricing algorithms, automation, and the daily handoffs where operations break.
Side Projects
Live systems. Not mockups.
ARC, Solomon
Multi-agent adversarial fund architecture. An AI council of scouts, analysts, and a risk sentinel with veto power debates capital allocation while deterministic math manages risk. Anthropic vs OpenAI, same architecture, different brains, side by side.
AEGIS
Real-time OSINT dashboard. 3D globe with live event tracking from USGS, NOAA, NASA, GDELT, 13 RSS defense feeds, Polymarket, and financial data. Claude-powered analyst generates intelligence briefings every 4 minutes.
Thinking
Theory and architecture behind the systems.
The Solomon Origin
Why LLMs should generate signals while math manages risk. How adversarial multi-agent councils create better decisions than single-model prompts. The full architectural thesis behind an AI-native fund.
Read →Where AI Actually Helps a Service Business
Most service businesses don't have an AI problem. They have handoff, visibility, and fragmentation problems. Where agents earn their place between office, field, and customer, and where they don't.
Read →The Conductor Thesis
Systems thinking and problem abstraction, applied to building with AI. Before anything gets built: why is this a problem, who does it affect, what does success look like, and does it still hold as the technology changes. Not one-shot prompts. Pairing with AI through the problem, then the build.