Build real deal solutions, not just prompts.
Connect email, folders, Excel, databases and dashboards into one system your team runs on. AI does the heavy lifting. You keep the judgment and control.
Taught by Joshua SampsonSame deal. Two ways to work it.
Before
Forty files. Numbers re-keyed by hand. A tracker rebuilt every week. A model and memo that do not tie. A chat window that forgets.
After
One standardized workflow every deal runs through. Every number sourced. The model and memo tie out. A deal brain your team owns. Faster, and the same every time.
What stops most people, and how the build handles it.
I can't code.
Claude writes the code. You direct it, and a check tells you it works.
AI makes things up.
Every number carries its source, and the model and memo are audited.
My firm won't let me.
You build on your own machine, with sample deals made for the course.
I'll get stuck.
Every part ends with a check, plus a monthly live build session.
It won't connect to our tools.
Connecting them is the course: email, folders, Excel, SQL and APIs.
Taught by Joshua Sampson
Joshua has spent over 7 years building solutions across commercial real estate, structured finance, and Corporate and Investment Banking, including the data and analytics platforms behind institutional investment teams at NYMT, Barings, and Wells Fargo. That combination, deal judgment and the systems underneath it, is the firm.
Previously New York Mortgage Trust · Barings · Wells Fargo
*Underwritten at New York Mortgage Trust. Deal value reflects the total capitalization of deals analyzed. Data and analytics platforms built at New York Mortgage Trust, Barings, and Wells Fargo.
Notes
*Underwritten at New York Mortgage Trust. Deal value reflects the total capitalization of deals analyzed. Data and analytics platforms built at New York Mortgage Trust, Barings, and Wells Fargo.
The deal room keeps itself.
Rent roll and T-12, cleaned up.
The market, read for you.
What you get.
The build: one practice deal from the broker's email to investment committee.
See all the parts
- Part 0Toolbox
- Part 1Deal record
- Part 2Prospect screen
- Part 3Deal room
- Part 4Package reads
- Part 5Research page
- Part 6Underwriting model
- Part 7Investment memo
- Part 8Pipeline
- CapstoneSecond deal
Every part is a solution on its own: use one piece or the whole system.
A check at the end of every part that proves it works.
A monthly live build session, recorded.
A searchable index of every concept, linked to the exact spot in the videos.
Everything you build is yours: your repo, your database, your workflow.
Start free: the Automated Market Report class.
Join the waitlist.
One email when it opens, and occasional notes on the build. See our Privacy Policy.
Questions
- How do I learn AI for commercial real estate?
- By building something you will use. Here you build a Deal Workspace with Claude and run a practice deal through it, from the broker's email to the investment committee memo.
- How does it work?
- Self-paced, plus a monthly group session that builds one piece live. You build on your own machine, in your own repo and database, and every part ends with a check that tells you whether it works.
- Do I need to know how to code?
- No. Claude writes the code. You decide what it builds, and a check tells you whether it works.
- What tools do I need?
- A paid Claude plan that includes Claude Code, plus GitHub, a free Supabase project, Python, Excel and, for market data, your own Esri key or free Census data. These tools are not ours, and some cost money.
- Can I use my firm's deals?
- Not in the course. You build on your own machine with sample deals built for it.
- When does it open, and what does it cost?
- Neither is set yet. Join the waitlist and you will hear first.