DealNgn — the platform behind STAX Real Estate

Sourcing to closing.
One system.

Commercial brokerage is nine disconnected tools and a person holding it together. DealNgn is the whole continuum as software — finding the asset, identifying the owner, underwriting it, winning the listing, marketing it, matching the buyer, papering the contract, closing the deal.

It is not a demo, a pilot, or a pitch-deck roadmap. It runs a licensed brokerage's live book every day — built by people who spent a decade shipping enterprise software before they ever carried a real estate license, and who have since opened and operated 43 of these businesses themselves.
Start an investor conversation → See what it actually does
dealngn.com / command-center
P1
Callback — owner, 4 stations
Asked for BOV · 2 days elapsed
09:15
P1
Inspection period expires
Contract deadline · 3 days out
T−3
P2
Re-engage — dormant 241 days
Surfaced by cadence rule
Today
DONE
Offering package + data room ready
Overnight · no human touch
04:22
Representative surface · illustrative data
PRODUCTION INSTANCE — RUNNING NOW AUTONOMOUS AGENTS ACROSS 8 DEAL STAGES BUILT BY OPERATORS WHO WERE ENGINEERS FIRST
9,700+
Sites under
continuous coverage
660+
Sites underwritten
to one standard
$300M+
Live deal
pipeline
18
States
covered
$525M+
CRE sales
volume
250+
Stations & c-stores
sold or leased
Our work has been covered in
The Verge NAIOP Development Magazine Commercial Construction & Renovation Chain Store Age The Real Deal The Robot Report C-Store Dive Convenience Store News CStore Decisions
The ceiling

Commercial real estate still runs on stale data and manual work.

Not because the people are bad at it. Because the work does not scale — and everyone in the industry has quietly accepted the ceiling that creates.

01

The market you can see is the market that's listed.

The best assets in this class rarely reach a portal. If your view of the market is what's publicly for sale, you are shopping from the leftovers and calling it a market survey.

02

Underwriting is artisanal.

Every broker models a deal a little differently, so no two opinions of value are truly comparable. Fine on one deal. Across a hundred, nobody can tell you which one is actually the best use of capital.

03

A great broker can hold about forty deals in their head.

That is the real constraint on a brokerage — not talent, not capital. Everything past that number goes unwatched, and the follow-up that would have made the deal never happens.

The platform

Everyone else built a tool. We built the continuum.

The CRE software market is a shelf of point solutions — a data subscription here, a marketing tool there, a CRM nobody updates. DealNgn spans the entire transaction lifecycle in one system with one data model. We don't publish how it works. We'll happily show you what it spans.

01Source 02Research 03Underwrite 04Win listing 05Market 06Match 07Negotiate 08Close DEALNGN — ONE SYSTEM ONE DATA MODEL · NO EXPORT, NO RE-KEYING, NO SEAMS Data subscription Underwriting spreadsheet CRM — export, re-key, import Marketing / OM tool Contract / e-sign NO POINT SOLUTION COVERS THIS
Transaction stageDealNgnWhat else you would buy
STAGE 01
Source
Data subscription
STAGE 02
Research
Data subscription
STAGE 03
Underwrite
Underwriting spreadsheet
STAGE 04
Win the listing
CRM
STAGE 05
Market
Marketing / OM tool
STAGE 06
Match
CRM — after a re-key
STAGE 07
Negotiate
No point solution covers this
STAGE 08
Close
Contract / e-sign
ONE SYSTEM, ONE DATA MODEL NO EXPORT · NO RE-KEYING · NO SEAMS
STAGE 01
Source
  • Off-market asset discovery
  • Nationwide parcel enrichment
  • Chain / independent classification
  • Continuous re-scan for change events
7,900+ sites tracked
STAGE 02
Research
  • Owner entity resolution
  • Beneficial-ownership mapping
  • Reachability scoring
  • Demographics & traffic counts
Automated per-site dossier
STAGE 03
Underwrite
  • One standardized valuation model
  • Fuel volume & c-store margin logic
  • Expense benchmarking
  • Sensitivity + debt solve (DSCR / LTV)
660+ sites, one standard
STAGE 04
Win the listing
  • Multi-channel owner outreach
  • AI voice agent with warm transfer
  • Interest and timing captured on every call
  • BOV & proposal generation
Nothing goes cold
STAGE 05
Market
  • Offering memoranda, auto-composed
  • Confidential + on-market tracks
  • Data room with gated access
  • Portal & broadcast distribution
Institutional package on every deal
STAGE 06
Match
  • Structured buyer buy-box model
  • Multi-dimension match scoring
  • Geographic rule resolution
  • Daily surfacing of new fits
Ranked, not blasted
STAGE 07
Negotiate
  • Offer matrix & comparison
  • LOI and contract generation
  • NDA workflow with approval gates
  • Counterparty activity timeline
Multi-offer tracked side by side
STAGE 08
Close
  • Deadline watchdog & escalation
  • Title & estoppel review
  • Diligence document assembly
  • Closing file organization
Nothing missed on the calendar
EIGHT STAGES · ONE DATA MODEL NO EXPORT, NO RE-KEYING, NO SEAMS BETWEEN STAGES
Inside the system

What it looks like on a Tuesday.

Representative surfaces from the production platform. Deal names, owners and figures below are illustrative — the interfaces are real.

dealngn.com / command-center
P1
Callback — owner, 4 stations, Polk County
Asked for BOV · 2 days elapsed · heat 91
09:15
P1
Inspection period expires — Chevron / Winter Garden
Contract deadline · 3 days out
T−3
P2
New buyer inquiry — 1031, $4–7M, FL Gulf Coast
Auto-replied 5m · 3 listings matched
11:02
P2
Re-engage — "call me after the holidays" (Nov)
Dormant 241 days · surfaced by cadence rule
Today
DONE
OM generated + data room provisioned
Overnight · no human touch
04:22
Command Center. Every open thread in the brokerage, priority-ranked before anyone sits down. The forty-deal ceiling, removed.
dealngn.com / property / 4417 · underwriting
Branded Fuel + C-Store
Hillsborough County, FL
4 MPD · 3,100 SF · AADT 28,400 · fee simple
A−
Value
$4.62M
Cap
7.15%
DSCR
1.48×
CONSERVATIVE $4.10M
BASE $4.62M
BEST CASE $5.18M
Standardized underwriting. Fuel volume, c-store margin, opex benchmarks, environmental status and debt terms resolved by one model — so a hundred assets can actually be ranked against each other.
dealngn.com / listing / 218 · buyer match
94
Family office — Southeast, all-cash
$3–8M · FL required · branded preferred · fee simple
VIPNDA
88
1031 exchange — 41 days remaining
$4–6M · NNN only · no operations · Gulf Coast
HOT
76
Operator — 11 stations, expanding north
$2–5M · will take operations · Tampa metro
OP
71
Private fund — Tier 2 markets, levered
$3–12M · 7%+ cap · excl. environmental open
FUND
Buyer matching. Every buyer's criteria is structured data, not a note in someone's inbox. A new listing is ranked against the whole book in seconds — and the reverse runs nightly.
dealngn.com / coverage
ME AK VT NH WA ID MT ND MN WI MI NY MA OR NV WY SD IA IL IN OH PA NJ CA UT CO NE MO KY WV VA MD CT AZ NM KS AR TN NC SC DE RI OK LA MS AL GA HI TX FL
Primary market Established Coverage layer
Florida
Primary market · full-depth enrichment
Texas
Metro-level buy-box resolution
Georgia
Metro-level buy-box resolution
Tennessee · Nevada
Active transaction markets
+ 13 additional states
Coverage layer live · enrichment expanding
Coverage, not listings. We watch assets whether or not the owner has decided to sell. By the time something is officially for sale, we have usually been tracking it for a year.
The voice layer

It is not a demo.
It is on the phone.

Owner outreach is the hardest, least scalable, most avoided work in brokerage. The voice layer does the first mile of it — dials the owner, has a real conversation, handles the brush-off, and warm-transfers a live human seller to a licensed broker who can actually help them.

  • Outbound and inbound. It prospects, and it answers.
  • Warm transfer, not a callback form. An interested owner is connected to a licensed broker on the same call.
  • Everything lands in the record. Transcript, objection class, next action and follow-up date written to the deal automatically.
  • It never decides anything. The voice layer opens conversations. A licensed human makes every judgment that matters.
Outbound → owner, 2 stations, Central FL 02:41
AI

Hi — is this the owner of the station on Orange Blossom Trail? I'm calling from STAX. We work almost exclusively on fuel and c-store properties in Florida.

Owner

I'm not selling.

AI

Understood — most of the owners we talk to aren't. Can I ask one thing before I let you go? We've been tracking your site, and given where fuel volumes and cap rates sit right now, our read is that it's worth meaningfully more than most owners in your area assume. Would it be useful to just know the number?

Owner

…What are you seeing on something like mine?

AI

Let me put you straight through to Michael — he's our principal broker, and he has operated these himself, not just brokered them. One moment.

WARM TRANSFER → LICENSED BROKER · LOGGED TO DEAL RECORD

Illustrative transcript. Composite of real call patterns; identifying details removed.

The part everyone gets wrong

This is not a robot that presses play.

DealNgn doesn't replace judgment. It arms it. Every deal that moves is a decision a licensed broker made — with better information than anyone else in the room had.

The platform does the work that scales: watching, screening, underwriting, remembering. The work that doesn't scale stays human — reading a seller across a kitchen table, structuring a sale-leaseback around what a family actually needs, knowing when a number on the page is wrong. That division is the entire design, and it is the reason this works on real transactions instead of in a demo.

Anyone promising a brokerage that runs itself is selling you something that will eventually cost a deal and possibly a license. Leverage is the product. The rest is a story.

A first-year agent on this platform operates with the market awareness of a fifteen-year veteran. A fifteen-year veteran operates like a team of ten.
Michael Salafia, founder and principal broker of STAX Real Estate
Who built it

A decade shipping enterprise software. Forty-three gas station businesses opened and operated. A principal broker's license.

Almost nobody has all three, and you need all three to know what to build. Every feature here earned its place by surviving a transaction — because the people who wrote it are the ones who lose the listing when it is wrong.

Why this team, and why this asset class →
For investors

We built it for ourselves.
The question is what it's worth to everyone else.

We are not raising publicly and this page is not an offering. We are, however, in conversations with venture investors and family offices who see what we see: that the same platform running one specialist brokerage could run a category — and that the team who built it is the rarest configuration in vertical AI — enterprise software careers, a real operating history in the asset class, and the licenses to transact in it.

We will show you considerably more in a room than we will on a public page, and you will understand immediately why that is. The fastest path is a direct one — you'll be talking to the founder, not a corp dev associate.

Start a conversation → See the brokerage
What a first conversation covers
  • Live walkthrough of the production system on the real book
  • Unit economics — cost per underwrite, per listing, per closing
  • The expansion thesis — asset classes, geographies, sequencing
  • Architecture — multi-tenancy, data moat, what scales and what doesn't
  • Where we're honest about risk — key-person, regulatory, competitive
For agents

The best reason to join a brokerage is the unfair advantage it hands you.

You still have to be good on the phone, and you still have to earn trust with an owner who has been pitched by six brokers this year. But you will walk into those conversations knowing more about their asset than they expect anyone to know — and you will never lose a deal because it fell off your list.

See open roles →