Tyler Sellars, CEO, Cactus
 Underwriting CRE at Speed with AI
Subscribe now on any of these platforms so you don't miss a single episode.
Apple Podcasts
Blubrry
Android
YouTube
TuneIn
Stitcher
Google
Spotify
Amazon Music
RSS
Guest: Tyler Sellars, CEO, Cactus
In brief:
-
Accuracy in underwriting matters more than speed alone.
-
Manual data handling increases error risk as deal volume rises.
-
Automation shifts underwriting from data entry to decision-making.
-
Underwriting volume, not capital, is a primary acquisition constraint.
-
Sponsors prefer tools that integrate with, rather than replace, Excel models.
Why acquisitions are a bottleneck
Â
In today's acquisition environment, capital is rarely the limiting factor (when done properly i.e. with systems like those we provide clients). Underwriting capacity is.
Most commercial real estate teams still rely on workflows built around manual data transfer. Offering memoranda, rent rolls, and T12s arrive in fragmented formats, requiring analysts to re-enter and reconcile information before analysis can begin. Assumptions are layered manually. Market data is pulled separately.
Â
As Tyler Sellars, co-founder and CEO of Cactus, explains, "In Excel or Argus, it takes time to be able to implement the data coming in from PDF or Excel files that you're receiving from the data rooms in an acquisition or development deal."
Â
Speed without accuracy creates false confidence
Â
Many investors have experimented with generic AI tools - like ChatGPT, Claude, Gemini etc. - to accelerate underwriting but results have been mixed, largely because speed is useless without reliability and accuracy.
Â
Yet underwriting errors rarely reveal themselves immediately. They surface later, often in missed assumptions, credibility loss, or capital misallocation. Cactus addresses this by emphasizing accuracy and on-screen auditability (a really cool part of the platform's offering). Every extracted data point can be traced back to its source document, with visual confirmation showing exactly where the information originated.
Â
Data rooms as a source of hidden risk
Â
Something you've probably noticed is how often deal documents contradict each other i.e. the pitch deck/offering memo doesn't tally precisely with the offering documents, for example or how rent assumptions in an OM may not match the rent roll, vacancy figures differing across files, loss-to-lease assumptions may be inconsistent etc.
Â
Rather than treating documents independently, Cactus audits across them simultaneously allowing discrepancies to surface immediately.
Â
Market comps auto-loaded
In the workflow Tyler demonstrated, rental comps, concessions, unit mixes, and similarity scores, taken directly from sources you likely already use. are integrated directly into underwriting output, together with ownership information and rental histories provide additional context.
Â
Without having to jump around from app to app or website to website, Cactus brings in the comps you need from the sources you already use automatically, saving time, increasingly efficiency, and helping you screen deals in a tiny fraction of the time you currently spend.
Â
Excel remains central
What I love about Cactus is that it does not force users to abandon Excel.
Â
"Your Excel is part of your IP," Sellars says. Instead, the platform maps extracted data, comps, and outputs directly into your existing pro-forma/underwriting models, preserving formulas and structure.
Â
You can also use Cactus's own underwriting model - and should probably test it. It's not going to be as sophisticated as your own, proprietary in-house underwriting models but it's advanced and will help surface opportunities that warrant full analysis very quickly.
Â
What's really cool is that, while you can create a Cactus template that houses and auto-populates all your major screening assumptions to come to a conclusion about a deal, you can also upload your own Excel model and have Cactus populate that with data-room numbers too.
Â
Stress testing without Excel
Cactus also simplifies scenario analysis. Users can adjust purchase price, financing terms, or hold period in a sandbox environment without overwriting original assumptions and watch results update instantly.
Â
This allows acquisitions teams to test scenarios live without doing full analyses, further accelerating the deal screening process.
Â
Bottom line
The Cactus tool will help you underwrite vastly more opportunities than you are currently managing while increasing speed and accuracy.
Â
For investors wanting higher deal velocity (is there anyone who doesn't), Cactus is a distinct advantage in a highly competitive environment.
Related to this episode:
