Neuhaus Realty Group is where it all started. Austin residential real estate — buying, selling, investing — since 2009. Full-service brokerage, TREC #593057.
But it’s not a typical brokerage website. Over the years, I’ve turned neuhausre.com into one of the most AI-integrated real estate platforms in Austin. Most agents use AI to write listing descriptions. I built an entire platform with it — custom WordPress theme and plugin, MLS-integrated property search, automated valuations, and 500+ pages of location-specific content generated and maintained through AI pipelines.
What it does
The Virtual Office Website (VOW) gives clients full MLS property search — active and under-contract listings with more detail than a standard IDX feed. Saved searches, property alerts, lead capture, and gated access to sold data for serious buyers. The search queries a PostgreSQL-backed property database that syncs with MLS data, serving results with sub-second response times.
Automated valuations powered by the Valuation Engine — the self-improving comp-based prediction system that runs nightly experiments to improve its accuracy. CMA reports and broker price opinions pull from the same engine, combining algorithmic analysis with the manual adjustments that make an appraisal defensible.
500+ location pages covering Austin neighborhoods, school comparisons, and market analysis. Each page targets specific search queries with real data — median home prices, school ratings, neighborhood characteristics, recent sales activity. The content is generated through AI pipelines and maintained as market data updates, so a page about Lake Travis real estate reflects this month’s numbers, not last year’s.
Lead generation through gated resources — downloadable guides, STR investment analysis, home valuation tools. Each capture point routes through the FUB MCP Server into Follow Up Boss with automatic tagging, stage assignment, and action plan enrollment.
How it’s built
The platform is WordPress with a custom theme and a custom plugin, both built from scratch with Claude. The theme handles all front-end rendering — property search UI, location pages, blog templates, lead capture forms, structured data markup. The plugin handles the backend: MLS data synchronization, VOW authentication, property search queries, sitemap generation, and integrations with external services.
The property data layer runs on a dedicated database server. MLS listings sync through a data pipeline, with the property search API serving queries through an application layer. The VOW public API — which also powers the Local Austin MCP — exposes active listings through a rate-limited, Cloudflare-protected endpoint. Sold data stays behind VOW authentication as required by MLS rules.
The content pipeline is the most AI-intensive part of the operation. Generating and maintaining 500+ location pages, school comparisons, neighborhood guides, and market analysis posts requires a systematic approach. Each page goes through generation, fact-checking, SEO optimization, and publishing stages. When market data changes — new median prices, updated school ratings, zoning changes — the pipeline regenerates affected sections. A separate fact-checking system verifies claims against authoritative sources before content goes live.
SEO is engineered into the platform architecture, not applied as an afterthought. Automated sitemap generation handles the thousands of URLs the VOW creates. Canonical URL management prevents duplicate content across property pages. JSON-LD structured data (RealEstateListing, Article, Person, LocalBusiness schemas) is built into the theme templates. Server-side caching keeps page loads fast despite the dynamic WordPress layer.
The tool ecosystem ties everything together. The FUB MCP Server manages all CRM operations through natural language — contact updates, deal management, task assignment, follow-up sequences. The Local Austin MCP provides instant property and civic data lookups across three counties. The Valuation Engine runs autonomous nightly experiments to improve price predictions. Each tool was built to solve a specific friction point in the brokerage workflow, and together they form an integrated AI-assisted operation that competes with portals backed by engineering teams.
Why I built it
I started Neuhaus Realty Group in 2009 because I wanted to do real estate differently — more data, less BS, direct communication. The technology side evolved naturally. Every time I hit a limitation with off-the-shelf tools, I built something better. The IDX feed was too limited, so I built a VOW. The CRM was too clicky, so I built an MCP server. The valuations were too manual, so I built a self-improving engine.
Eventually, the building became as interesting as the selling. This brokerage is where I learned that AI could be a genuine business partner, not just a toy. It’s the proving ground for every other project on this page.