I underwrite rental housing and build AI tools for that work.

I’m Herman Chan, Director of Acquisitions at Haven Realty Capital, based in the San Francisco Bay Area. My work covers sourcing, underwriting and market research for build-to-rent, single-family rental and multifamily investments across Sunbelt markets.

From 2011 to 2020, I owned and operated single-family rental homes in Houston and Phoenix, handling acquisitions, financing, renovations and property management.

I started using AI in late February 2026, with no prior engineering background. I began building tools alongside my acquisitions work, starting with the tasks I kept repeating: reading deal documents, organizing financial statements, comparing rents and researching competing development.

That work became the three systems on this site: an underwriting agent, a supply database and a rent database. Each grew out of a specific part of my own workflow.

I use AI to write the code. My role is to define the workflow, decide how the pieces fit together and specify what a correct result should look like. I review the implementation and test the outputs against their sources. When something fails, I work through it with AI and verify the fix. I also operate the systems and review the analyses before using them in an investment decision.

This site brings those projects together with sample outputs and short notes on the build process: the decisions I made, the failures I investigated and the changes that followed. It is a record of what I’m building and learning while doing the work.

I hold a Master of Real Estate Development from the University of Southern California and studied management science at UC San Diego.

I’m also a father, and outside work I enjoy running and training for HYROX.