/ INVESTOR
1 / 13
YYZDATA · INVESTOR

Two engines. One company.

Forge, an applied-AI services practice that pays the bills today. Signals, an ambient intelligence platform for physical space that compounds into the company.
Seed raise · $1M · Q4 2026
Pre-seed closed Nov 2025
Confidential
The problem

Physical businesses run blind.

The data that decides a good night already exists. It is thrown away the moment it happens.
Cameras
Record everything, say nothing
Surveillance video is written to disk, reviewed only after something goes wrong.
Labor
Staffed by guesswork
Hospitality labor runs 25 to 35 percent of revenue and climbing, scheduled against a hunch, not the actual crowd.
Queues
Walkouts are invisible
45 to 80 percent of customers abandon a purchase when a line looks long. Nobody sees it until the register misses.
Systems
Four tools, no picture
POS, reservations, labor and cameras never talk. The operator is the integration layer.
Labor: National Restaurant Association, 2026. Abandonment: Oracle concessions survey; MLB fan study, Apr 2025.
The product

Signals: cameras in, answers out.

An on-site intelligence layer over the cameras a venue already runs. Aggregate counts and events by default: no faces, no identity, no video stored. Alerts reach a phone before a problem becomes a loss.
Live today
Working system
Two-camera prototype: on-device detection, live dashboard with per-role views, WhatsApp alerts including proactive understaffing warnings.
Live today
Model pipeline
Detector fine-tuned on real, licensed venue footage. Synthetic data generator and training loop proven end to end.
Live today
Connectors
POS, reservations and labor feed the same layer; revenue sits next to foot traffic on one screen.
Queue and wait intelligence and historical patterns: in build, 90 to 180 days.
Proof, honestly labeled

Built, not promised.

Live
The prototype
Real cameras, real detection, real alerts, running now, with a detector trained on real venue footage.
In motion
First build partner
A DC lounge, warm relationship, design partnership. First install and first before-and-after numbers.
Pipeline
Lined up behind it
A university system in conversation, multiple lounge pilots in discussion, and a warm channel into Fenway Park. Pipeline, not booked revenue, and we say so.
Demo
Vertical walkthroughs
Interactive demos for stadium, lounge, dealership and other venue types. Illustrative, not deployments.
The market

The near ground is empty, and countable.

70,015 US bars and nightclubs, roughly 15,000 licensed dispensaries, roughly 7,700 cigar and hookah lounges: about 93,000 venues in the segments we can serve today. At our blended $200 per month that is roughly $220M of serviceable ARR, before restaurants (749,000) or large venues.
93K
Near-segment US venues
~$220M
Serviceable ARR at list
~420
Venues to $1M ARR
$12.7B → $38B
Video analytics, 2024 to 2030
IBISWorld 2026 (bars); Flowhub 2026 (dispensaries); industry profile (lounges); Grand View Research (market, 19.5% CAGR, edge fastest at 26.3%).
Why now

Three doors just opened.

Regulation
Privacy is a moat now
FTC actions and BIPA-class laws punish identity surveillance. Aggregate-only by design skips the entire fight, and incumbents cannot follow without abandoning their feature lists.
Economics
Edge got cheap
An Apache-licensed detector on a $300 box, no cloud inference, no video storage. A cost floor cloud-era incumbents cannot reach.
Incumbents
Their revenue is our wedge
Solink ($175/mo, video-first, 18,000+ locations) and Verkada (per-camera hardware and licenses) monetize storage and gear. Riding a customer's existing cameras cannibalizes them. It is our whole model.
Go-to-market

Land small. Prove fast. Expand.

Now
Build partner
The DC lounge: deploy, learn install hours, capture the first quotable KPI under a signed data license.
Next
Camera-mandated retail
Venues legally required to run cameras already have the hardware and the budget line. Software-only, fastest close.
Then
Bars and restaurants
The volume market, entered through POS connectors and reference stories.
Held warm
Large venues
Referral-driven conversations only. Zero custom builds until the small-venue motion prints references.
Business model

Software margins, priced for an owner.

$99 to $199 per month on a venue's own cameras, where COGS is close to zero and gross margin approaches software-pure. $249 to $599 with a managed kit; the setup fee recovers hardware in one to two months. Roughly 83 percent gross margin at the entry kit tier, with blended margin at scale targeting the mid-80s. No capital decision anywhere in the sale.
83%
Gross margin, entry tier
1–2 mo
Hardware payback via setup fee
$9.7K
3-yr gross per mid-tier venue
$0
Customer CapEx
Engine two

Forge: the practice that funds the platform.

An applied-intelligence services arm with real clients and real delivery. ISO 27001 and ISO 9001 certified. Leadership has founded, run and advised businesses from startup to large enterprise, and built regulated data systems in finance and life sciences along the way. Owner, operator, customer, partner, builder: every side of the table.
Studio
Custom builds
Custom models, applied vision systems, intelligent workflows and agents, platform build and delivery. Engagements from $20K assessments to $150K+ platforms, fixed-price and phased.
Express
Packaged SMB builds
Storefront and commerce, POS and operations, automation, web presence. Direct integrations, not platform fees.
Run
Production support
Monthly retainers: monitoring, support, model ops. Recurring services revenue.
Inside Signals
The delivery muscle
Every install is a Forge engagement: setup and integration fees keep deployment off the product P&L. Enterprise rollouts are Forge programs.
We know commoditization first-hand: our earliest offers included synthetic data and model training, and foundation models swallowed them within a year. We shifted early. The rule that survives every model release: start from the business problem, not the technology.
The moat

The demo is the door. The data is the moat.

Any single feature can be cloned. The compounding asset cannot:
Flywheel
Licensed venue data
Every site signs an explicit, paid-for training license. Every venue improves the model the next venue buys. Incumbents' contracts and stored-video liabilities make this hard to retrofit.
Architecture
Privacy-native, edge-native
Aggregate-only from the first commit, processing on site. This is a design, not a setting, and it clears compliance review in one conversation.
Layer
The integration seam
Video plus POS, reservations, labor and sensors in one reasoning layer. The more systems connected, the harder we are to remove.
The vision

Every signal. One console.

Cameras are only the first signal, because they are the richest one a business already produces. The same layer is built to ingest anything that describes how an operation actually runs: IoT and sensors, transactions, documents and text, sound. Prevention, prediction, forecasting and execution from one console.
Philosophy
The system bends to the business
AI removes the old excuse for forcing businesses to change their process to fit software. We build to your process, your rules, your vocabulary. Best practices are suggested, never imposed.
Consequence
Verticals are configurations
A lounge, a dispensary, a showroom and a concourse are the same problem at different scales. One platform, retargeted by connector and persona, not rebuilt per industry.
Endgame
The operating layer
Not another ERP, POS or CRM. The intelligence on top of all of them: their systems stay the system of record, ours becomes the system of action.
Use of funds

Build. Deploy. Prove.

We are not searching for the product. We are scaling its delivery.
~40% · Product
Finish the platform's first mile
Queue and wait intelligence, historical patterns, the connector suite (Toast, Square, Dutchie), drop-and-deploy kit hardened to zero-touch.
~35% · Build partners & test sites
10 to 15 instrumented venues
The DC lounge first, more lounges and camera-mandated retail behind it, a university system in conversation. Compliant hardware stock, install economics measured site by site, data licenses signed at every one.
~25% · Traction
Turn test sites into proof
Paid conversions, the referral channel activated, one quotable ROI number per vertical: the evidence pack Series A is priced on.
The ask

$1M seed, closing Q4 2026.

The $205K pre-seed, from a small and strategic cap table of operators and executives, bought the product, the pivot and the lessons. Services revenue is coming in, cash is conserved, and the founders have drawn nothing to date. The seed budgets modest founder sustenance alongside the build, deliberately: companies die of running out of cash, and of running out of founder.
$1M
Seed, SAFE
$205K
Pre-seed closed Nov 2025
30–50
Venues live on seed capital
Q3–Q4 '27
Series A target
hello@yyzdata.com
YYZDATA · ACCESS CODE