Building AI systems that connect knowledge, evidence, and action.
I work at the intersection of applied AI, organizational intelligence, and empirical research. I'm particularly interested in systems that can reason over complex bodies of evidence, recognize what they don't know, and translate reliable conclusions into useful work.
My background spans startup operations, technology law, and experimental psychology. I've led business and legal operations at venture-backed technology companies, practiced as a corporate attorney, and conducted doctoral research at Oxford. Increasingly, my work focuses on designing and building AI systems that solve problems I've encountered firsthand.
Ladder — From organizational memory to operational intelligence
A company intelligence system that ingests institutional knowledge, evaluates operational maturity, identifies evidence-backed gaps, and helps close them.
- Memory: A searchable, cited knowledge layer spanning company documents and records, accessible through MCP.
- Maturity: An operational reference model covering 83 areas across 11 company functions, with evidence-grounded assessments and explicit treatment of missing information.
- Leverage: Automation opportunities qualified against actual organizational readiness, rather than assumed capabilities.
Includes a working maturity assessment board and a document-generation pipeline with citation validation, independent quality review, and human approval. The knowledge and diagnostic layers have been exercised against real company environments; live operational automation remains the next stage of development.
Mrkt — Moneyball for transactional law
An empirical research pipeline investigating whether negotiated provisions in M&A agreements contain measurable information about subsequent deal outcomes.
Uses LLM-based structured extraction, SEC-filed agreements, financial market data, and statistical analysis. The initial research covers 152 merger agreements, with validation against expert annotations and multiple robustness tests.
An experiment in what becomes possible when legal domain expertise, machine-assisted analysis, and quantitative research meet.
Charlie — Autonomous intelligence and adversarial evaluation
A multi-agent research system built to follow structural change in the entertainment industry.
Combines source ingestion, analytical agents, daily editorial synthesis, persistent thesis development, and an independent adversarial reviewer that challenges unsupported claims and weak inference.
Designed for ongoing use, with scheduled execution, human feedback, and explicit separation between incoming evidence and evolving interpretation.
TBC Intelligence — Client-built business development intelligence
A business-development intelligence system built for and used by Tre Borden /Co, combining LLM-assisted prospect research with structured municipal building-permit data.
Uses deterministic opportunity scoring across Los Angeles, New York, and San Francisco to identify potential art commissioning opportunities for a creative business.
The interesting problems in applied AI often begin where the demonstration ends.
- Evidence before confidence. Distinguish missing facts from missing access to facts.
- Provenance matters. Preserve the distinction between original evidence, model inference, and generated output.
- Evaluate the system, not just the answer. Build validation, adversarial review, and measurable failure modes into the workflow.
- Automation should be earned. Make prerequisites explicit, preserve human decision rights, and avoid automating processes that aren't ready.
I favor practical architectures, inspectable assumptions, and systems whose limitations are as clear as their capabilities.
Oxford DPhil, Experimental Psychology · Yale Law School JD
Former technology attorney and startup operations executive, with experience spanning venture-backed infrastructure companies, corporate transactions, compliance, and organizational operations.
My current interests include enterprise knowledge systems, agentic workflows, AI evaluation, applied inference, and the organizational design required to deploy AI responsibly.
Interested in collaborating on applied AI, company intelligence, and technically ambitious early-stage businesses.