Building an AI-native private capital firm in Texas.
Uplift Capital — multifamily real estate private equity. Operating partner. The person who both structures the capital and ships the code that runs it.
About
From institutional capital to AI-native operating.
A decade in institutional private equity managing portfolios north of $20 billion. Today, operating partner at Uplift Capital — a Texas multifamily real estate private equity firm. AI-native builder of the systems that run it.
I built Uplift OS, a small portfolio of agent-run operating companies, and the public research tools below — largely solo. The point was to find out how far one operator fluent in both capital and code could get with an agent harness. Far enough that the next step is a team where that combination compounds.
Read more →Selected work
Systems I've built
Production systems, not slides. Each one shipped solo, against real operating data.
Datacenter Registry
Built a global datacenter registry with per-field provenance — facilities, power, chips, cooling, and fiber modeled as a graph — plus a fully static map dashboard exposing the physical infrastructure behind the AI buildout.
Read case studySpotWire
Built a market-intelligence engine that ingests surplus, channel, and public data across the AI buildout supply chain and publishes four weekly indices — compute, interconnect, power, and facility — each a per-component rebased composite with every input cited to a source trust tier and a versioned methodology.
Read case studyPressure-Testing a Treasury Strategy to Death
A systematic Treasury-futures curve strategy backtested at a respectable Sharpe over 33 years — until honest 3-leg trade booking revealed most of the 'edge' was an accounting artifact and the strategy quietly lost to T-bills. The interesting result isn't the strategy; it's the discipline that killed it.
Read case studyBoxscore
Built a local-first Rust harness that explains why a property's NOI missed or beat budget — computing a deterministic revenue-expense-NOI bridge, ranking the GL accounts driving the gap, and writing an evidence-backed, plain-English variance narrative with every claim traced to its source row.
Read case studyGridX Civic Intelligence OS
Built a local-first intelligence system that turns public-source briefs, permit portals, RFP notices, and infrastructure signals into a structured, searchable archive of civic and corridor intelligence across the North Texas / Texoma region.
Read case studyMultifamily Underwriting Engine
Built a deterministic underwriting engine that ingests offering memorandums, rent rolls, and T12 financials and produces full month-by-month cashflow, IRR, DSCR, and exit-cap analysis — with every output reconciled against an institutional model and traceable back to its source documents.
Read case studyCommunity Bank Value Screener
Built a fully auditable screener that ranks U.S. listed community banks on price-to-tangible-book, capital strength, credit quality, and insider alignment — turning a 247-name universe into a short list of 129 names that clear seven hard gates, each traced to FFIEC Call Reports and SEC filings.
Read case studyMarket Study Agent
Built an agent-driven platform that produces institutional-quality multifamily market studies, combining public data pipelines with AI analysis and compliance-first reporting.
Read case studyUplift OS
Built the operational backbone for a $140M multifamily real estate portfolio — a full-stack dashboard ingesting rent rolls, comp analysis, turnover reports, and 500K+ general ledger transactions into real-time unit economics and asset management intelligence.
Read case studyRedevelopment Intelligence System
Built a data-driven system that identifies parcels with rising redevelopment pressure by analyzing the relationship between land value, improvement value, and time using public tax records and GIS.
Read case studyNotes
Recent writing
Practitioner essays on capital, code, and operating.
June 20, 2026
Markets Run on Stories. Can You Measure One?
A research note on treating market narratives as contagions — borrowing the math of epidemics to model how a story forms, accelerates, peaks, and collapses. What the framework computes, what's real, and what's still just a hypothesis.
Read →June 8, 2026
What I Mean by AI-Native
Most AI-for-business content describes tools that make humans faster. AI-native is a different structural claim: agents handle the recurring operating layer, humans concentrate on judgment. What this looks like across the businesses I actually run.
Read →May 17, 2026
I Used AI to Build a Small-Cap Bank Screener. It Runs Without AI.
How Claude Code and Codex pair-programmed a systematic screening pipeline across five public databases — and why the most honest thing about the system is that the AI built the machine, not the thesis.
Read →Working together
Open to the right seat.
Most of my time belongs to Uplift and the portfolio. But I built that operating stack solo to prove a point: one person fluent in both capital and code, armed with an agent harness, can do the work that used to need a team. The next step in leverage is a team where that compounds rather than bottlenecks on one person.
Full-time or contract, the test is the same — is this a hard, operational, data-rich problem where capital judgment and working software have to live in the same head? If that's what you're building, I want to hear about it.
Start a conversation →The Letter
Compound
Bi-weekly signed research on Texas multifamily, REITs, and private-capital flows. Free archive. Paid tier coming.
Subscribe to Compound →