Cannabis retail intelligence

See next week’s demand before it arrives.

Dispensight combines Bayesian forecasting, real-time KPI analytics, and fraud detection — purpose-built for cannabis retail operators.

HMC–NUTS engine · TechPOS native · built for BC cannabis retail

Weekly demand forecast
Per store · per SKU
highlow nowforecast →← actuals
Actual unitsForecast80% credible interval
10%
Avg daily error
80%
Calibrated intervals
4
MCMC chains
~10%
Average daily forecast error, live deployment
80%
Honest, well-calibrated prediction intervals
92%+
SecureLeaf fraud-detection accuracy
8
Store locations tracked in real time
Features

Built for the complexity of cannabis retail.

Every feature is shaped by how BC dispensaries actually run — the data sources, the regulations, and the day-to-day.

Bayesian Forecasting

NUTS/HMC sampler running four chains across every KPI. Probabilistic predictions with uncertainty intervals — not false-precision point estimates.

Real-Time KPI Tracking

Near-real-time dashboards for basket size, retention, team performance, and revenue across up to eight store locations.

Security-First Architecture

CSRF protection, IP-based brute-force guards, SSRF mitigations, role-based access control, and full audit logging throughout.

BCLDB Compliance

BI tuned for BC cannabis regulation: shift reconciliation, return monitoring, and discount tracking built to match the rules you operate under.

External Signal Integration

Forecasts fold in weather, local events, and other external signals — so you can see sales swings coming before they land.

TechPOS Native

Ingests CSV exports from your existing POS. No middleware, no manual wrangling — just upload and analyze.

Platform preview

See Dispensight in action.

A live look at the dashboards your team will actually use — from real-time intraday tracking to ML-powered 7-day forecasts.

How it works

From raw POS export to a forecast you can act on.

Four steps, and only the first one needs you.

01

Connect your data

Drop a TechPOS CSV export — no middleware, no manual wrangling. Setup is once.

02

The engine models demand

A NUTS/HMC sampler fits demand per store and SKU and quantifies the uncertainty.

03

You get a forecast

Fan charts and prediction intervals show the likely range for the week ahead, not false precision.

04

Trust builds over time

Every forecast is scored against what actually happened, in the open.

Validation & transparency

A forecast is only worth trusting if it’s checked.

Dispensight publishes its accuracy openly — real deployments, out-of-sample validation, and honest interval calibration. See the numbers for yourself in the research below.

  • Forecasts scored against actuals — error, coverage, and bias reported, not hidden.
  • Out-of-sample validation studies you can read end to end.
  • Security-first architecture with full audit logging throughout the stack.
Engine accuracy · live deployment
Validated
Average daily error~10%
80% interval coveragewell-calibrated
SamplerNUTS/HMC · 4 chains
Fraud API accuracy92%+
Validation reportpublished
Research

Evidence-based cannabis retail intelligence.

Dispensight is backed by published work — forecasting validation, Bayesian methodology, and cross-industry CRM applied to dispensary operations.

🎯Validation

Forecasting Engine Accuracy

Real-world results from a live deployment: ~10% average daily error and honest, well-calibrated 80% prediction intervals.

Read
Validation

Basket Validation Study

Forecast vs actual — rigorous out-of-sample validation of the Dispensight forecasting engine.

Read
📈Bayesian

Basket Theta MCMC

Full Bayesian forecast validation using MCMC sampling on real basket-size data.

Read
⚙️Bayesian

Bayesian Engines · Convergence Study

Comparing sampler convergence and diagnostic properties across model configurations.

Read
📊Validation

Average Basket Forecast & Validation

A forecast-and-validation walkthrough on average basket size using the theta MCMC report.

Read
🛩️CRM

Crew Resource Management

From aviation cockpit to cannabis dispensary — a cross-industry CRM framework.

Read
✈️CRM

The One-Seat Fallacy

Why single-employee dependency is costing dispensaries more than they realize.

Read
📘CRM

Breaking the Turnover Cycle

CRM principles applied to cannabis retail — retaining staff and customers at once.

Read
🧠CRM

Budtender Brain Drain

Mapping talent-loss patterns and their downstream effects on dispensary revenue.

Read
🎓Theory

Human Learning & Bayesianism

Connecting cognitive science to how retailers update beliefs about customer behavior.

Read
🗺️Strategy

Standardize or Localize?

Practical implications for multi-location cannabis retail strategy and brand consistency.

Read
📍Regional

Regional vs Local Strategies

Data-driven analysis of when regional standardization outperforms local adaptation.

Read

Ready to see Dispensight in action?

Launch the app to explore your own stores, or book a walkthrough and we’ll show you what it sees in your data.