Breaking Property News 11/8/26
Daily bite-sized proptech and property news in partnership with Proptech-X.
Why the ‘Deal File’ is the hidden layer behind AI in Commercial Real Estate
The invisible world of ‘worked judgement’ that all AI-generated decks sit on
Todd Terry is the co-founder and CTO of Ascendix Technologies. Over the years, Todd has designed and delivered solutions for many thousands of users from Fortune 500 companies in financial services and commercial real estate to a variety of small and mid-market B2B enterprises. Along with enterprise CRM solutions, he has also delivered innovative software products leveraging technologies in cloud computing, big data, natural language search and cross-platform mobility.
In a recent thought leadership piece, Todd articulates the position that, ‘AI promised commercial real estate professionals it would automate the deck, the memo, and the pitch. It delivered. But every high-stakes deck sits on a body of work you cannot see. The preparation, the objections, the alternatives, the sourcing. AI can make that body of work easier to draft and easier to keep, yet too often the reasoning still disappears after the deck is done.’ Todd calls that that body of work the deal file. And he goes on to explain what it is, and why AI-generated decks often fall apart without it, and how to prevent this from happening.
Todd further explains that the deal file is the body of worked judgment behind every important commercial real estate deliverable. It contains the alternatives considered, the risks assessed, the objections anticipated, the recommendation logic and the sources behind the claims. A deck may be the visible output, but the deal file is what makes that output defensible when someone starts asking difficult questions. It is the preparation that sits underneath the recommendation and gives the team confidence that the conclusion can withstand scrutiny.
This distinction has become increasingly important as artificial intelligence becomes embedded in commercial real estate workflows. For several years, AI development has concentrated heavily on making the final deliverable faster and more attractive. Generate the slides. Format the memo. Improve the language. Produce the recommendation. The technology has become remarkably good at this part of the process.
But the deliverable was never the hardest part of high-stakes commercial real estate work. The difficult part is the thinking behind it.
Under an investment committee presentation, for example, there may be a detailed comparable set and the reasoning behind why particular properties were selected. There may be downside scenarios that never appeared on a slide, alternative assets that were considered and rejected, assumptions that required testing and objections that the team expected the investment committee to raise.
That is the worked judgment that makes the final recommendation meaningful. Without it, an impressive presentation can quickly fall apart under questioning.
Every serious CRE deliverable can effectively be viewed as having three layers.
The first is context: leases, rent rolls, emails, CRM records, market information and the other source material that provides the raw information.
The second is the deal file, where that information is turned into worked judgment.
The third is the trust envelope, which carries the evidence behind the finished deliverable and allows a reader to understand which claims are sourced, derived, assumed or unsupported.
These are different jobs. When they are all thrown into a single folder or system, the organisation may retain the raw information and the finished presentation while losing the most valuable layer in between: the reasoning that connects the two.
The opportunity presented by generative AI is that this middle layer is now far cheaper to create and maintain. The alternatives considered, downside cases, objection maps, risk registers and recommendation rationale can all be documented as part of the normal workflow. What previously required substantial analyst time can increasingly be captured while the work is being done.
That changes the economics of the deal file, as instead of treating the preparation as disposable work that exists only to support one meeting, organisations can retain it as an asset. The deal file can become a reusable record of how a decision was reached, why alternatives were rejected and which assumptions mattered.
Yet many organisations still allow this reasoning to disappear. The deck is presented and the file is saved and the preparation behind it gradually evaporates into someone’s memory or remains buried in an AI chat that nobody will ever open again. The next transaction then starts with a blank page, even when the organisation has already solved many of the same problems before.
Keeping the deal file changes that
The first benefit is defensibility. When a senior decision-maker asks where an assumption came from, the answer does not depend on someone’s memory. The supporting reasoning is already documented. If an investment committee challenges an exit-cap assumption, the team can identify where it came from and understand how it affects the recommendation.
The second benefit is reuse. The next transaction does not have to start from scratch. Comparable-property logic, risk frameworks, previous objections and the reasoning behind earlier decisions can provide a foundation for the next analysis. The new deal can then be tested against that accumulated knowledge rather than forcing the team to recreate it.
The third benefit is retention of institutional knowledge. A deal file locked inside the head of a senior analyst leaves when that person leaves the organisation. A deal file retained as a shared business asset remains available to the next person who has to make a similar decision. Over time, those individual files become a record of the firm’s accumulated judgment.
This is where the real opportunity for AI in commercial real estate may lie. The objective should not simply be to make the deck faster, prettier or easier to produce. It should be to preserve the thinking that makes the deck valuable in the first place.
A strong AI-enabled workflow should assemble the right context, turn that context into a durable deal file and carry the supporting evidence through to the final deliverable. The system should provide the structure and preserve the work, while the commercial real estate professional provides the expertise, judgment and final decision.
This also points to an important change in how AI systems should be designed. The structure of a deal file should not depend on one particularly enthusiastic power user creating prompts and spreadsheets. The scaffolding should already exist within the workflow. The professional should not have to remember what needs to be preserved every time a new transaction begins.
The result is more than a better presentation. It is a more intelligent and increasingly institutionalised decision-making process. The next time a meeting ends, and the finished presentation is filed away, the important question is therefore not simply whether the deck has been saved. It is whether everything that made the deck credible has been saved with it.
Because in commercial real estate, the deck may be what people see, but the deal file is where the judgment lives.
Andrew Stanton Executive Editor – moving property and proptech forward. PropTech-X
You May Also Enjoy
The summer improvements that could add value to your house
Turning anxiety into advantage: how SMEs can make AI work for them
Home gym can add more than £24,500 to the value of a new-build home
Beachfront homes command £1.4m as Britain’s love affair with the seaside endures
Four to five houses is the sweet spot for first-time buyers
