02Data Lab
Applied experiments in dashboards, machine learning, and AI. Each project will ship with documentation of its data sources, assumptions, evaluation, and known limitations.
Status
The Housing Affordability Explorer is live in this preview. All other projects below are in development and shown as placeholders.Explore how home prices, mortgage rates, down payments, and household income shape the cost of buying a home. Compare two purchase scenarios with adjustable assumptions and transparent calculations.
Companion tool to The Price of Staying: Washington, DC — Research Working Paper 01.
Placeholder
A scenario tool for comparing the long-run costs of renting and owning.
Planned: transparent assumptions for rates, appreciation, maintenance, and opportunity cost that users can adjust.
Placeholder
A documented, explainable model for estimating residential list prices.
Planned: feature engineering from property and neighborhood attributes, with model cards describing data, validation, and known failure modes.
Placeholder
An experiment in forecasting listing duration by submarket.
Planned: a time-series approach evaluated against simple baselines, with uncertainty intervals and honest error reporting.
Placeholder
An AI assistant that drafts plain-language neighborhood market briefs from public data.
Planned: retrieval over documented sources with citations on every claim, and clear labeling of AI-generated text.
Placeholder
An AI tool to help agents and clients navigate lengthy property disclosures.
Planned: summarization with source highlighting and guardrails that defer to licensed professionals.
Model cards
Every model will document its training data, validation approach, and failure modes.
Honest baselines
Models are compared against simple baselines, and errors are reported plainly.
Labeled AI
AI-generated content will be labeled, cited, and never presented as professional advice.