DealDesk/Features/Comps Library

Comps + scoring

AI-scored comparables, with reasoning.

Save any abstract as a comp in one click. Score similar deals by reasoning, not keywords. Median $/SF + lease-term distribution by submarket.

What it does

01

One-click save-as-comp

Every completed abstract can become a comp. Three-tier reveal model controls what teammates and counterparties can see.

02

Similarity scoring with bullets

For any anchor comp, we surface 8 nearby same-submarket comps with a 0-100 match score and 5 reasoning bullets per match. Hover for the model’s logic.

03

Submarket trends

Median $/SF, total SF moved, lease-term distribution by submarket — the data brokers pitch tenants with.

How it works

From cold start to shipped result.

  1. 01 — Import or save

    Save abstracts as comps; manual entry; or CSV import. PostGIS-indexed for radius search.

  2. 02 — Filter and score

    Six-facet filter rail with URL-state-sync. Save filter combinations as named views.

  3. 03 — Surface for a deal

    On any comp page, the Similar Comps rail shows the eight strongest matches with reasoning.

Why brokers love it

I stopped asking the team for comps in Slack. The library actually has them now.

Director, tenant-rep — New York

8

scored matches per anchor

0-100

match score with reasoning

3

reveal tiers (private/market/public)

Free for the first three abstracts every month.

No credit card. No sign-up to try the abstractor. The output is yours to keep.