The product

Under three seconds. Zero extra labor. Every bottle.

The mechanism and the proof in the same motion. A patented sensing mat reads each bottle in under three seconds with no caps, no scales, and no disruption to service. The cloud AI tracks, detects, predicts, and reconciles. The detail that matters to an investor is not the sensor. It is that the model requires zero extra labor, and that every venue feeds an AI no new entrant can match.

under 3s
Per-bottle read

Hybrid RFID and optical.

0.1 fl oz
Read accuracy

[pending source verification]

8
POS systems

Reconciled against the point of sale.

6s
Pour to cloud

Real-time, not batch.

The hardware

The sensing mat

The mat sits under the bottles on the back bar and reads each one using hybrid RFID and optical detection. Hardware-enabled, not software-only, which is what makes the read objective and automatic rather than a human judgment call. There are no spout caps, no kitchen scales, and no change to how a bartender works. The bottle goes back on the mat, the same as always, and it has been measured.

No caps

Nothing attaches to the spout. Free-pour service is untouched.

No scales

No bottle is lifted onto a connected scale by hand.

No disruption

The mat is passive. Service runs exactly as it did before.

The mat carries a one-time installation fee per venue, roughly $1,000, so hardware volume scales with venue count rather than usage. That makes the supply requirement plannable against the pipeline.

The intelligence

The cloud AI

Every read flows to a cloud AI engine that does four things continuously, for every venue.

FunctionWhat it does
TrackMaintains a live inventory of every bottle, updated as bottles return to the mat.
DetectFlags shrinkage and variance as it happens, not days later in a weekly count.
PredictForecasts demand and reorder timing per venue from its own pour history.
ReconcileMatches measured pours against the point of sale to surface unrecorded loss.

The cloud is the recurring product: roughly $399 per month per venue, the line that makes this an Inventory-as-a-Service business rather than a hardware purchase.

The surface

KlynkzLive

KlynkzLive is the app the manager actually lives in: real-time inventory, shrinkage alerts, and reorder guidance on any device. It is the reason managers do not cancel. A tool that surfaces a live, trustworthy number, where there used to be a stale weekly snapshot, becomes part of the daily workflow. Stickiness comes from that workflow embedding plus the ROI, not from a contract lock.

The integrations

Reconciled against eight POS systems

Klynkz reconciles measured pours against the point of sale, which is where unrecorded loss hides. It integrates with eight POS systems, the same systems the natural strategic acquirers already own, which is also why that embedding is itself a moat.

Oracle MICROS
Aloha POS
Toast
Clover
Square
Lightspeed
Revel
NCR Aloha
The killer detail

Zero extra labor is the model, not a feature

Bottles return to the mat and inventory updates itself. There is no count to run, no scale to load, no app to fill in. The labor cost of inventory does not shrink. It disappears.

This is the structural difference from every alternative. BevSpot and WISK still require a human to capture data by photo or entry. Smart scales need a person to place each bottle. Flow meters cover only metered lines and alter service. Klynkz is the only system that is automatic, real-time, and free-pour at once, which is a genuine white-space position, not a feature checkbox.

The business fact

The AI flywheel

The flywheel is a business fact, not a slogan. Every venue feeds the AI, and the data set compounds into accuracy a new entrant cannot match without installing thousands of mats and waiting years. This is the bridge to the moat.

The flywheel
More venuesSmarter AIHigher marginsMore venues
The supervised learning loop
  1. 01
    Data capture

    Every pour logged: bottle, brand, volume, time, location, workstation.

  2. 02
    Prediction

    The AI predicts demand, reorder timing, and shrinkage risk per venue.

  3. 03
    Real-world test

    Predictions are compared to actuals: ordered versus consumed versus wasted.

  4. 04
    Model refinement

    The gap between prediction and reality trains more accurate models network-wide.

33,000+
Bottles in database

[pending source verification]

550,000+
Transactions processed

[pending source verification]

6s
Pour to intelligence

Network-wide learning.

The network effect is the point: every new venue makes the AI smarter for all venues. Real-world performance compounds rather than degrades, and the data set becomes the asset a competitor cannot reconstruct. That is where the moat page picks up.