Reef

Reef pulls OMs, Excel files, ARGUS runs, rent rolls, financials, broker assumptions, and support documents into one intake workflow. Nobody re-keys numbers from a PDF into a spreadsheet by hand. Starting from a single intake record means the rest of the platform works from the same data instead of a scatter of versions.

From the extracted deal record, Reef builds a structured Excel model automatically. It's a populated model, not a blank template, and it runs inside your firm's approved structure once you upload that template. The point is speed without abandoning how your team already underwrites.

Each deal is grounded against market benchmarks and property tax data for the relevant US market. The platform flags rents that trade materially above or below the benchmark, so the model reflects market reality rather than only the sponsor's narrative. This is where a deal gets a defensible baseline.

Reef surfaces issues before committee questions appear. It catches tenant concentration, occupancy dependence, rent anomalies, and cross-document inconsistencies as the package comes together. Seeing a single tenant holding 34% of net rentable area before you open the model is very different from finding it in a late review.

When the model is ready, Reef assembles a complete IC memo that synthesizes the deal's thesis, risks, market positioning, and financial summary. You edit it in a built-in rich text editor and apply your firm's branding. Before export, an AI proofreader scans for data inconsistencies, unsupported claims, and internal contradictions.

Firms upload their approved Excel templates and Reef generates within those standards. You can swap templates on a past deal to compare assumptions, and every template change is tracked with version history. That keeps firm logic, naming, and tabs intact across deals instead of forcing a one-size-fits-all model.

Every deal, assumption, and IC decision is stored in a searchable knowledge graph. Teams can find comparable deals by market, asset class, vintage, or sponsor, and answer questions like what cap rates they used in a given market two years ago. Prior deals compound judgment instead of disappearing into folders.

A side-by-side workspace shows the OM, the model, the benchmark, and the memo at once, so analysts stop juggling four windows. The "time machine" capability takes any past deal, applies current market benchmarks, and regenerates the model and memo with updated data, which helps with portfolio monitoring.