Case Study
The Structure an Attorney Would Paper, and the Business Logic They Do Not Provide
A value-add retail operator in Atlanta needed a syndication structure that was competitive for investors and safe for him. Here is what the system produced, and what his attorney was never going to give him.
Case Study
A system we built and ran. Client details are anonymized where required. Numbers are what we measured, not what a vendor reported.
"Attorneys aren't business people. They don't know the deal. They're just helping you be safe. This gave me good insight on how to set up the most competitive yet safe structure for my syndication."
Josh Ahlzadeh, Delza Properties
Josh buys and operates value-add retail inside about sixty miles of Atlanta. Lean shop, owner-operator, real deals. He was putting together a syndication and ran into the problem every first-time sponsor hits.
His attorney could tell him whether a structure was legal. His attorney could not tell him whether it was any good.
Those are different questions. One is about compliance. The other is about whether investors will actually fund the deal and whether the sponsor survives if it underperforms. Counsel papers what you decide. Deciding is somebody else's job, and for most small sponsors that somebody does not exist.
What the system produced
Three decisions, which together are the economic spine of a syndication.
Preferred return: 8 to 10%, cumulative. Paid to investors before the operator earns anything. The range was not pulled from memory. The system checked where value-add retail syndications were actually pricing in 2026 and confirmed the range was competitive before putting a number in front of him.
Distribution waterfall: tiered. Splits that step up as the deal performs, so the operator is rewarded hardest exactly where value-add upside shows. This is the difference between a structure that merely pays the sponsor and one that aligns the sponsor with the outcome investors care about.
GP gets paid last. Investors clear their preferred return and their capital before the operator participates. This is the single clearest signal to an LP that a sponsor is confident, and it is the thing a sponsor is most tempted to negotiate away.
How it got there
Four things happened that a generic model session would not have produced.
It already knew the business. The system carries a working memory of the operator: value-add retail, a specific geography, a lean owner-operator structure, his actual deals. It started from his situation rather than from a blank page, which is why the output was a structure rather than an explainer.
It checked the market before proposing a number. Preferred return ranges drift. Proposing 8 to 10% because it appeared in training data would have been a guess dressed as advice. It verified against current comparable terms first.
It reasoned across the real frameworks together. LP/GP structure, the tiered IRR waterfall, the full sponsor fee stack, and the Regulation D 506(b) versus 506(c) fork. Those interact. A waterfall that looks generous means something different depending on where the fees sit and who you are allowed to solicit.
It pressure-tested for his downside. Promote structured non-recourse, a modest sponsor co-invest, and the loan guarantee explicitly flagged as the real personal exposure. The brief was competitive and safe, and the second half is the part that usually goes unexamined until it matters.
Why this is the pattern, not a one-off
The reusable lesson is not "AI can design a waterfall."
It is that the expensive professionals around a small operator are each solving a narrow, defensible slice. The attorney handles legal risk. The accountant handles tax treatment. The broker handles the transaction. Nobody is paid to hold the whole business question, and at Josh's scale nobody is on staff to do it either.
That gap is where a system with real memory of the operator earns its keep. Not by replacing counsel, which it cannot and should not, but by producing the informed position that makes counsel's work faster and the sponsor's decisions better.
Josh put it more directly than we would have: attorneys are helping you be safe. Safe is necessary. It is not a strategy.
The honest limits
This produced a structure to take to a lawyer. It did not paper the deal, it did not replace the securities work, and it did not remove the personal exposure on the loan guarantee.
It also worked because the system already knew his business. Dropped cold into a generic chat window, the same questions would have produced a competent-sounding overview of syndication structures and nothing decision-grade. The memory is the product. The reasoning is what the memory makes possible.
Sources
- 1Josh Ahlzadeh, Delza Properties, Atlanta. On-camera testimonial recorded 2026-06, used with permission. Delza acquires and operates value-add retail in the Atlanta market.
- 2Structure figures below are the actual output of the engagement, cross-checked against the working session transcript: 8-10% cumulative preferred return, tiered distribution waterfall, GP paid after LP preferred return and return of capital.
- 3Market calibration: the system checked 2026 comparable syndication terms for value-add retail before proposing a preferred return range, rather than relying on training data.
- 4Securities frameworks considered in the working session: LP/GP structure, tiered IRR waterfall, sponsor fee stack, and the Regulation D 506(b) versus 506(c) distinction.
- 5Disclosure: Delza Properties is a Woodworks Realty Studio client. This is our own work, published with the client's permission, and should be read as a case study rather than as independent reporting.
- 6Nothing here is legal, tax, or investment advice. Syndication structures are securities matters and require qualified counsel.
Keep reading
Work With Us
Your team already has AI. Almost nobody uses it.
The licences are bought and the accounts are open, and most of them sit in a tab nobody opens on a Tuesday. Woodworks Realty Studio gets an assistant into the hands of every person on a team, one at a time, each one set up around how that person actually works. My own business runs on it before any of it reaches anyone else.