Patterson Research works across quantitative finance and artificial intelligence. Founded by Charles Patterson, the practice investigates investment ideas and builds software for research, evaluation, and decision-making.
Software for testing quantitative claims, examining investment theses, and making research easier to inspect.
Reconstructs trading claims and tests their evidence against multiple testing, fragile parameters, and unseen data. Its verdicts distinguish a rejected claim from one the available data cannot resolve.
A market and portfolio research interface with linked charts, scanners, saved views, and replayable snapshots. Source timestamps and freshness checks help a reader understand what each displayed number represents.
The workspace brings together market observations, paper portfolio exposures, and completed-study excerpts. Recorded snapshots preserve the data and context behind a view so it can be inspected later.
Breaks a thesis into falsifiable claims, locks its evidence requirements, and records attempts to refute it. LLMs propose interpretations; deterministic checks enforce the record and constrain the resulting memo.
The workflow separates claim decomposition, evidence review, and verdicts. Unresolved claims remain visible, and example fixtures are labeled separately from live research records.
A research knowledge workflow built on GBrain. It connects source passages, attributed claims, research candidates, and decisions, with reviewed records available to AI agents through deterministic retrieval.
Human-readable, version-controlled records preserve evidence locators and limitations. Source assertions and unvalidated candidates keep distinct states, so retrieving an idea does not silently turn it into an established finding.
How advances in drug discovery connect to supplier economics, manufacturing, and the expectations already reflected in stock prices.
A framework for finding early evidence, testing an investment thesis, and comparing a compelling story with what the price requires.
Follow the writing on Patterson Research’s Substack.
Visit the publicationSystematic strategies, portfolio construction, and market structure. Research accounts for the search behind a result, execution costs, and the limits of the sample.
Agent workflows, evaluation pipelines, and research software. Model outputs enter systems with explicit checks, evidence requirements, and review points.
Technology, industry economics, and valuation. Investigations connect a proposed mechanism to observable evidence and ask what would invalidate the thesis.
Researcher & software engineer
I founded Patterson Research to bring software engineering and investment research into the same practice. I build the tools, investigate the ideas, and work on making the evidence easier to question.
My background includes software engineering at Capital One and Potomac Fund Management, alongside independent quantitative research and trading infrastructure. The projects here reflect my interests in markets, AI systems, and empirical testing.
Available for select consulting engagements in AI systems, evaluation pipelines, and research tooling. Get in touch to discuss a project or compare research notes.
Consulting & research collaboration