Research, analysis, and perspectives on the built world.
AI agents differ from chatbots and copilots in one critical way: they execute multi-step workflows using tools, memory, and planning. This piece explains how agents work, where institutional development teams are deploying them in 2026 -- site sourcing, due diligence, market analysis, IC prep -- and what to evaluate before selecting one.
April 25, 2026 ·4 min read
Foundation model selection has a direct impact on CRE workflow accuracy, cost and analytical depth. This guide maps GPT-4o, Claude, Gemini and o-series models to specific development tasks and provides an evaluation framework covering context window, tool use reliability, structured output fidelity and cost at scale. Teams routing by task type rather than defaulting to a single model cut inference cost 40-60% while improving accuracy on high-stakes outputs.
April 23, 2026 ·4 min read
Proptech and agentic AI are regularly grouped under the same umbrella but are not the same thing. This post explains what each category actually does, how their deployment models differ, where they fit together in an institutional development team's technology stack, and how to evaluate each correctly.
April 21, 2026 ·5 min read
Retrieval-augmented generation (RAG) is the architecture behind AI tools that can answer questions about your documents accurately. This guide explains how it works, why it matters for CRE document workflows, and what to evaluate when selecting RAG-based platforms for lease abstraction, title review, and due diligence.
April 19, 2026 ·6 min read
AI agent orchestration -- the coordination of multiple AI calls, tool uses, and decision points -- enables complex multi-step development workflows no single prompt can complete. This post explains the three components of orchestrated AI (tools, memory, planning), where orchestration is deployed in real estate development today (site screening, due diligence, IC prep, permit tracking), and honest limitations to evaluate before buying. A practical guide for development teams assessing AI platforms beyond the chatbot layer.
April 18, 2026 ·5 min read
AI deployment in CRE development stalls not because the technology fails, but because teams sequence it wrong. This guide covers where to start, how to prepare data, whether to build or buy, and what to measure to prove ROI across the development lifecycle.
April 16, 2026 ·4 min read
A practical guide for enterprise CRE development teams on deploying AI at scale, covering which workflows to start with, how to sequence implementation, build-vs-buy decisions and what metrics to track. Written for CDOs and SVPs of Development evaluating how to move from pilots to portfolio-wide deployment.
April 15, 2026 ·4 min read
Most AI implementations in CRE development fail not because the technology does not work but because teams start in the wrong place. This guide covers how to sequence deployment, what data readiness looks like, and what to measure in the first year.
April 14, 2026 ·5 min read
Multimodal AI — models that process images, maps, and documents alongside text — extends AI capabilities into the visual layer of real estate development: site plans, surveys, construction photos, and zoning maps. This post covers the highest-value use cases in development workflows today, accuracy considerations for production deployment, and the implementation pattern that works.
April 12, 2026 ·5 min read