Work with models

Model a project end to end. Level 1 answers the delivery question — will it finish on time and on budget — and comes in two flavours: a waterfall schedule network simulated task by task, or an agile delivery flow where a backlog is drained by a throughput resampled from the team’s own history. Either one can start from a worked example that the AI turns into a model, or from your own MS-Project plan or Jira export.

Level 2 asks the harder question: will it deliver the value it was approved for. Thirty-two standard objectives with their targets, gates and stakeholders, producing the probability of project success — so schedule and cost become two objectives among many, not the whole answer.

Level 3 consolidates those same projects into a corporate control layer across entities: registers, period-by-period tracking of forecast against actuals, exception reporting and calibration. It never re-simulates.

Measure operational risk from your own loss history. Load a loss-event register — or one of the three examples, a bank’s, a manufacturer’s and a cyber and IT incident log — and each event type, at any Basel level or category, gets a frequency (Poisson or negative binomial) and a severity (lognormal, gamma, Weibull or empirical) fitted and tested against its own data. The simulation convolves them year by year, correlates the event types and returns expected loss, OpVaR, expected shortfall and the capital of each event type, with S-curves, a risk horizon, AI analysis and a PDF report. It runs in your browser.

Price the risk of a whole credit portfolio, loan by loan. Load your loan tape — or the 5,000-loan demo — and each credit gets a PD from its delinquency state through a transition matrix, a PERT LGD from its guarantee and a Bernoulli default, while segment-level PD, LGD and exposure move together with the cycle. The simulation charges funding and operating costs, reads economic capital off the loss distribution and returns expected and unexpected loss, RAROC, P(RAROC > hurdle) and EVA, for the whole book and by any column in your file. It runs in your browser.

Forecast the value of an investment portfolio in thousands of scenarios. Pick one of twelve model portfolios — or load your own price history and positions — and each asset is calibrated on its own prices while the assets move together through their historical correlation. The simulation returns the portfolio's value day by day with its fan chart, Value at Risk, Expected Shortfall, probability of loss and Sharpe ratio. It runs in your browser.

Measure liquidity risk from the daily balances of any number of banks. Load demand, savings and term deposits and liquid assets — or the five-bank demo — and every series is calibrated on its own history and calendar. The simulation returns each bank’s Liquidity-at-Risk against its buffer, core deposits, survival horizon and stress runs, with a backtest and Miller–Orr cash limits. It runs in your browser.

Turn a spreadsheet you already trust into a probabilistic one. Upload a finished deterministic model and iziRisk parses the real formulas on the server — not the last cached numbers — then picks the decision output, walks the dependency graph backwards to the hard-coded inputs that actually drive it, and ranks them with a ±10% tornado so you work on what matters. For each one it recommends an iz* distribution with its parameters and the reasoning behind them; you review, override anything you disagree with, and download a new .xlsm where every fixed input has become a distribution function and the rest of your formulas are untouched. Open it in Excel with the iziRisk Quantum add-in and simulate.

A growing catalogue of ready-made Excel workbooks you can download and run today. Each one is a complete quantitative risk model built on the iziRisk Quantum ribbon — distribution functions already wired to the uncertain inputs, Monte Carlo simulation, sensitivity analysis and other QRA techniques — covering cost estimates, schedules, business cases, contingency reserves and operational risk. Open one, replace the inputs with your own numbers, and you have a working model without starting from a blank sheet. Macro-enabled workbooks: enable the ribbon once and every model in the library behaves the same way.

Start from a blank sheet and a sentence. Describe a business problem in plain language — a bid, a budget, a contingency reserve, a make-or-buy decision — and the AI interviews you about what is uncertain and what you are actually deciding, then builds the workbook: the calculation framework, a distribution function on each uncertain input with the reasoning for choosing it, and the outputs worth watching. It runs the simulation and returns charts, a written interpretation and a presentation to download. You keep the Excel file, so the model is yours to adapt, audit and re-run.