iziRisk Projects · Level 1 · schedule and cost
Native CPM + Monte Carlo engine — the model a desktop MS-Project add-in will feed. Durations and lags are in working days (per the work-week calendar); dependencies support the four link types with lag (e.g. D:SS+5, C:FF-2), plus task constraints and deadlines. Edit the tasks, costs and risk register, then simulate.
Project and targets
Model and display
This sends content to an AI model (Anthropic). What is sent and how it is handled →
Tasks & durations
| Name | Calendar | Preds | min · ml · max | dist | Constraint | Deadline | |
|---|---|---|---|---|---|---|---|
| A | |||||||
| B | |||||||
| D | |||||||
| C | |||||||
| E |
Calendars (1)
Durations count working days of the task’s own calendar: a task on a 7-day calendar works through weekends. Holidays are days off; worked exception days are days worked although their weekday is not (a Saturday worked to recover time).
No extra calendars: every task works on the project calendar.
Probabilistic branches
Where the project can go one of several ways — a test passes or fails, a permit comes back clean or with conditions — name the paths, how likely each is, and the tasks that only happen on that path. Every iteration takes exactly one path; tasks on the others do no work, cost nothing, and their risks cannot fire. Link the last task of each path to the task where they rejoin.
No branches: the network runs every task in every iteration.
Cost elements
| Task | Fixed (min·ml·max) | Var/day (min·ml·max) | |
|---|---|---|---|
Joint confidence level — target convergence
Pick a probability of meeting BOTH the finish date and the budget together. iziRisk iteratively raises both targets over the last simulation’s joint distribution until that joint confidence is reached — giving you the date and budget to commit.
AI reports
Six AI agents, one for each reader. Four write from the last run — technical, governance, executive and plain language — the mitigation report from the run’s pricing of each response strategy, and the schedule-quality report from the task network itself. The checklist shows what the reports will draw on, and what is missing or out of date, before you spend an AI run. Every report stays in its own tab, with its PDF and its PowerPoint.
This sends content to an AI model (Anthropic). What is sent and how it is handled →
For the project controls lead and the risk analyst: every figure, the drivers, criticality and dependence.
For the PMO, the steering committee and audit: assumptions and evidence against GAO and AACE practice, the date and budget to commit to, numbered findings.
For the sponsor and management: on time, on budget, how likely, and what to do.
For anyone with a stake in the project, no project training needed: everyday words, no jargon.
For the risk owners: which response strategies to adopt, which are trade-offs, which are not worth their price.
For the planner: a 14-point logic check of the network (DCMA-14 style) and what to fix before trusting the forecast.