Decision Intelligence
Also: DI · decision intelligence software
A discipline that augments human decision-making with structured analysis, historical data, simulation and outcome tracking. Decision intelligence software does not replace the human decision-maker — it stress-tests the reasoning before a decision is locked in, then tracks what actually happened afterwards.
Decision intelligence combines elements of data science, behavioural economics and systems thinking to make business decisions more consistent and auditable. Unlike business intelligence (which tells you what happened) or AI assistants (which generate text), a decision intelligence system applies structured opposition, risk scoring and outcome loops to a specific decision at the moment it is being made.
Kauzio is built around decision intelligence as a core operating principle. Every decision submitted to Kauzio is scored for reversibility, stress-tested with a Monte Carlo simulation, argued from both sides, and then signed with a cryptographic receipt before being filed in the operator's decision history.
Decision Operating System
Also: Decision OS · DOS
A software layer that sits above a business's data and gives every decision a consistent, auditable process — from opposition and simulation through to a signed record and a 30/60/90/365-day outcome check-in. Kauzio Pulse is a decision operating system built for retail and hospitality operators.
The term "operating system" in this context is deliberate. Just as a computer OS standardises how every program interacts with hardware, a Decision OS standardises how every business decision interacts with data, safeguards and institutional memory.
A decision operating system does not own the decision. The operator does. What it does is ensure every decision — from a price change to a lease signing — goes through the same ten safeguards, gets recorded with the same level of detail, and is revisited at the same intervals. Over time, this creates a corpus of institutional decision knowledge that makes future calls sharper.
Sleep Lock
Also: sleep lock decision · decision sleep lock · cool-off period
A backend-enforced waiting period applied to high-risk or irreversible decisions. The verdict is server-side redacted until the lock expires. The idea is that the version of you that wakes up tomorrow — with a clear head and no urgency bias — gets the final say on irreversible calls.
Sleep locks are one of the ten Kauzio safeguards. They are not optional pop-up warnings. The decision verdict is cryptographically withheld at the server level until the cool-off period elapses.
An override is allowed, but it requires a written reason and is permanently logged in the decision audit trail. This means any future review can see not just the decision, but the fact that the operator overrode a sleep lock and why.
Lock duration scales with reversibility score. Easy-to-undo decisions get a short or no lock. One-way decisions — lease signings, permanent hires, public pricing — get longer locks by default.
Decision Receipt
Also: signed decision receipt · decision certificate · decision record
A cryptographically signed, timestamped record of a decision. It captures the question, both sides of the argument, the verdict, the reasoning, and the risk score. The signature cannot be altered after the fact — any modification breaks the hash.
A Kauzio decision receipt is more than a log entry. It is a tamper-evident artifact that proves what was decided, when, and what was argued. It is year-sequenced, HMAC-signed, and publicly verifiable at kauzio.com/verify/{certificate_number}.
Decision receipts matter for governance, investor relations, regulatory compliance, and institutional memory. When a pricing decision lands under scrutiny six months later, the receipt shows exactly what analysis was done and what both sides of the argument said — not a reconstructed memory of what happened.
Receipts are also outcome-extendable. As outcomes come in at 30, 60, 90 and 365 days, they are appended to the original signed record so the full history — prediction vs. reality — is preserved in one place.
Outcome Loop
Also: outcome tracking · decision outcome review · 30 90 365 review
An automated check-in system that fires at 30, 60, 90 and 365 days after a decision is made. It records what actually happened against what was predicted, grades the forecast, and feeds the result back into the calibration model so future decisions are sharper.
Most decision tools stop at the moment the decision is made. Kauzio does not. The outcome loop is what closes the learning cycle.
Check-ins are dispatched via Slack, Teams or email if the outcome cannot be auto-resolved from connected POS or accounting data. The operator rates the outcome, adds context, and the result is appended to the original signed decision receipt.
Over time, the outcome loop builds a per-operator corpus of decision performance data. This is what powers calibration: once enough outcomes are recorded, the system can show an operator their actual decision accuracy by category, identify where their judgment is systematically off, and adjust opposition weighting accordingly.
Calibration
Also: decision calibration · forecast calibration · accuracy calibration
The process by which Kauzio measures its own forecasting accuracy per sector, computes the gap between predictions and outcomes, and adjusts opposition weighting and sleep-lock thresholds to compensate. Calibration scores are published publicly.
Calibration in Kauzio is not a marketing term. It is a nightly computation: for every sector with at least 500 closed outcomes, Kauzio computes a calibration coefficient — the average ratio of predicted to actual outcomes across all decision types.
The coefficient feeds back into two places. First, opposition weighting: if a sector's decisions are systematically overconfident on a given dimension (e.g. margin impact), that dimension gets more weight in future opposition for that sector. Second, sleep-lock duration: sectors with lower calibration scores get longer default lock periods on irreversible decisions.
Calibration scores are public. Any potential customer can see exactly how well Kauzio's models perform in their sector before committing.
Monte Carlo Simulation
Also: Monte Carlo analysis · Monte Carlo business · stochastic simulation
A technique that runs thousands of randomised scenarios to produce a probability distribution of outcomes rather than a single-point forecast. Kauzio runs 1,000 Monte Carlo scenarios per decision across eight categories to estimate the range of possible results.
Monte Carlo simulation is standard in finance and engineering but rare in day-to-day business decision software. Kauzio applies it to every significant decision to produce an outcome distribution rather than a false-precision point estimate.
The 1,000 scenarios are stratified by decision type: hiring, expansion, marketing spend, staffing, contract signing, supplier switch, policy change, and generic. Pricing, stock orders and promotions have specialised Monte Carlo models that account for demand elasticity, competitor response and lead-time variance.
The result is not one number but a range with confidence intervals. An operator can see the optimistic case, the base case, and the downside case — and understand how wide the spread is. A narrow spread means the decision is robust to uncertainty. A wide spread means there is material risk that the base case will not materialise.
Reversibility Score
Also: decision reversibility · reversibility classification
A score from 1 to 3 that classifies every decision by how easy it is to undo: 1 = easy (change your mind at low cost), 2 = costly (undoable but with friction), 3 = irreversible (one-way door). Irreversible decisions trigger a longer sleep lock by default.
The reversibility score is one of the earliest signals Kauzio applies, because the reversibility of a decision should change how much scrutiny it gets before being locked in. A temporary promotion price is reversible. A lease is not. A test-hire with a probation period is more reversible than a senior permanent appointment.
Kauzio auto-classifies reversibility based on decision type, financial magnitude and time horizon. Operators can override the classification with a written reason, which is logged.
Reversibility is also used in post-decision review. When outcomes are collected, decisions that were marked irreversible get additional scrutiny — the system flags where irreversible bets performed poorly so the operator can adjust their risk appetite on that type of decision.
Causal Analysis
Also: root cause analysis · sales driver analysis · revenue attribution
The process of breaking down what actually drove a business metric up or down — isolating the percentage contribution of each factor (product mix, promotions, weather, seasonality, customer churn) rather than just observing that a number changed.
Most business dashboards tell you that sales dropped 12%. Causal analysis tells you that 6% of that drop came from a weather-driven footfall reduction, 4% from a supplier stock issue on your top SKU, and 2% from seasonal baseline decline — and that underlying customer repeat rate actually improved.
Kauzio's causal analysis module applies statistical attribution modelling to your transaction data. Each factor is scored for its contribution to the change, and each score carries a confidence interval. Where the data is thin, Kauzio says so rather than inventing a spurious attribution.
The output is a factor-by-factor breakdown with a causal patterns view over time — so you can see whether a given driver is a one-off (bad weather week) or structural (a product category in structural decline).
Signals (Monitor)
Also: business alerts · data alerts · anomaly detection
Automated alerts that fire when a business metric moves outside its expected range. Kauzio monitors connected data sources continuously and generates signals — tiered by severity — when something unusual happens, so operators know before it becomes a problem.
Traditional reporting shows you the past. Signals show you the present — specifically, the moments when the present is diverging from what history says it should look like.
Kauzio Signals uses statistical anomaly detection: for each metric (sales, margin, foot traffic, stock levels, return rate), Kauzio computes an expected range based on your historical data and seasonality. When actuals fall outside that range by more than the threshold, a signal fires.
Signals are tiered by severity: high-severity signals appear at the top of the dashboard immediately; medium and low signals are collected in the Monitor feed. From any signal, an operator can open a Kauzio decision directly — the signal context pre-fills the decision question, so the response is structured from the start.
Decision Playbooks
Also: decision templates · decision frameworks
Pre-built decision templates for common business calls — price changes, supplier switches, new hires, lease renewals, promotions. A playbook pre-configures the question structure, safeguards and risk criteria so operators start with the right framework, not a blank page.
The cost of a poor decision is often not the wrong answer — it is the failure to ask the right questions before committing. Decision playbooks exist to ensure that the right questions are always asked, regardless of who is making the call or how much time pressure they are under.
Kauzio ships with a standard playbook library covering the decision types that appear most often across retail and hospitality operators. Each playbook includes the decision question, key sub-questions, pre-configured safeguards (reversibility classification, risk scoring parameters, sleep-lock duration), and a suggested opposition checklist.
Account owners can create custom playbooks for decisions specific to their business — for example, a franchise with a standard lease renewal process can encode that process into a playbook that any manager in the network can apply consistently.
Team Bets (Prediction Markets)
Also: team prediction market · internal prediction market · business betting
A prediction market feature in Kauzio where team members bet virtual points on business outcomes. Aggregating predictions surfaces collective intelligence and identifies who reads the business most accurately over time.
Prediction markets have a long history in economics as mechanisms for aggregating dispersed information. The insight is that when people bet on an outcome, the aggregate of their predictions is often more accurate than any single forecast — including management's.
Kauzio Team Bets applies this to everyday business outcomes: will the promotion hit its revenue target? Will footfall be up or down next weekend? Will the new supplier deliver on time?
Team members bet with virtual points (no monetary value). Kauzio computes the implied probability from the betting spread. When the outcome resolves, points are settled and the leaderboard updates. Over time, the leaderboard identifies which individuals in the business have the most accurate intuition about specific outcome types — surfacing expertise that formal reporting structures often miss.
Decision Economics
Also: decision ROI · decision performance · expected vs actual value
A Kauzio module that tracks the expected versus actual financial value of decisions, grouped by decision type. It computes a return-on-investment figure per decision class and surfaces which types of calls consistently outperform or underperform their predictions.
Decision Economics is the closing loop of the decision intelligence cycle. After calibration has measured forecast accuracy and the outcome loop has recorded what happened, Decision Economics synthesises that data into a financial performance view by decision type.
For each category — pricing, hiring, promotions, supplier changes, expansions — Kauzio computes the total expected value (what the model predicted at decision time), the total actual value (what the outcome check-ins recorded), and the variance between them.
The variance tells an operator where their judgment is systematically biased. If pricing decisions consistently overperform expectations, the model should adjust upward. If promotion decisions consistently underperform, the optimism at decision time is a known bias to correct for.
This feedback loop is what makes Kauzio's models sharper over time — it is not a static tool. The more decisions are made and resolved, the more precisely the system can calibrate both its forecasts and its opposition.
Six-Axis Opposition
Also: decision opposition · structured opposition · AI opposition
A structured argument against a proposed decision across six dimensions: revenue risk, margin impact, demand elasticity, operational friction, reversal cost, and uncertainty. Kauzio generates the six-axis opposition using real customer data, not generic AI reasoning.
Most AI tools that claim to "challenge" a decision produce generic caveats that could apply to any decision. Kauzio's six-axis opposition is grounded in the operator's own historical data — past losses on the same type of decision, actual margin impacts from comparable calls, and real reversal costs from decisions that were unwound.
The six axes are:
1. Revenue risk — the downside range on top-line impact
2. Margin impact — how the decision affects gross and net margin
3. Demand elasticity — how price-sensitive the relevant customers or segments are
4. Operational friction — staff time, system changes and supplier coordination required
5. Reversal cost — what it costs to undo the decision if it goes wrong
6. Uncertainty — how much is unknown at the moment of the decision
Each axis is scored, and the composite score drives the overall opposition strength and the sleep-lock duration.
