projected annual value at steady state (Year 3+), risk-adjusted
projected 10-year gross value, probability-weighted
modeled full-stack payback on the build and run cost
second-order value vs. hard-dollar savings at steady state
01The modeled subject
The subject is a real firm, anonymized: a private real estate investment firm with $12B+ in assets under management, 200+ properties, roughly 600 employees, a $3.5B development pipeline, multiple non-traded REIT vehicles, and credit and adjacent platforms — run on an in-house legal team of about six attorneys. The firm publicly positions on speed of execution: closing challenging deals where others have failed.
That positioning is why the number is this large. The firm does not treat legal as overhead — legal review sits on the critical path of the thing it sells. Releasing roughly 980 in-house attorney hours a year and compressing document turnaround therefore prices as deal-throughput capacity, not as cost reduction. That reframe moves the buyer from the GC's expense line to the investment committee's revenue line, and it is where 84% of the projected value sits.
The platform modeled: a contract review agent wrapped in a deterministic harness — playbook-encoded negotiating positions, clause-level diff and risk flags, edit drafting in tracked changes, a measured error-rate gate per document type, and full audit logging. Attorneys verify; the harness earns skip-tier review only after the error gate clears on live volume.
The scenario table, the line-by-line impact model, the four-wave deployment plan, and the six dependencies that decide whether any of it materializes. Free, in exchange for a work email.