Platform
From your data to a decision your team owns
One platform carries every Anticipia product. It prepares forecasts and drafts from your data, checks them against your rules, and leaves every decision to a person.
From data to decision
Every product follows the same path, and each step hands its result to the next.
Your data
Admissions, ward stays, staff, skills and absences, or stock, usage and reports, come in as extracts you already produce. No live integration project comes first.
Engines
Forecasts, requirements, rosters and supply plans are computed as batch jobs. Each result carries the run, the software version and the model version that produced it.
Rules and access
Working-time rules, ratios and permissions are checked in code. Wards, shifts and rules are your configuration, with validity dates.
People decide
Drafts are reviewed, approved and published by people with the right to do so. Nothing is published automatically, and every change is logged.
Rules live in code
A language model can explain a result. It never decides whether a rule is kept. The checks that matter run as ordinary code, in one place, on every change.
Working-time rules
Rest between shifts, consecutive days and nights, absences and contracted hours are checked when you edit a roster and when the optimiser drafts one. The grid and the optimiser read the same rules.
Permissions on every action
Every action checks the permission it needs against its scope: a ward, a hospital or the whole organisation.
Settings, not code
Ratios, minimums, shift codes and rule profiles are versioned configuration. A new version starts from a date you choose, and the rules for past days never change.
Security and data protection
Separate for each organisation
Each organisation has its own databases, storage and login. The database also filters every row by organisation, as a second check beneath the application.
Pseudonymised patient data
Patient data is pseudonymised in the planning products. The assistant receives aggregates, never individual patient records.
Support access you grant
Our staff reach your data only through a grant that your administrators approve, with a level and an end time. You can revoke it at any moment, and every action is written to your audit log and to ours.
Real data stays out of development
Real patient-level data never goes into development, tests, logs or assistant prompts. Problems are reproduced with synthetic data of the same shape.
Your account, your roles
Staff sign in with their existing account. Groups in your identity provider map to roles, and an administrator sees and edits what each role may do.
An audit trail you can read
The audit screen lists who did what and when, filtered by person, date and grant, and exports to CSV.
Deployed where you decide
Infrastructure you control
Anticipia can run in your own cloud account or on your own servers. The same container images and deployment run in both.
One organisation per job
Every forecast, requirement and roster job runs for a single organisation, with only that organisation’s credentials. Secrets live in a secrets manager, never in configuration.
Your choice of model
- A cloud language model or a self-hosted one.
- An organisation that must not use cloud models never reaches one.
- The assistant keeps the same access whichever model runs.
Products you switch on
An organisation enables the products it wants. A product that is off is hidden, and a read-only product shows data but allows no changes.
What’s next
Where the platform goes next
The platform takes on more planning problems with the same forecasts, rules and approval steps. Two are next.
Theatre list scheduling
Next is bringing theatre list scheduling onto the same platform, with a person approving every list.
Staff simulation
Next is testing a change before making it: alter sickness, demand, closed beds or a staffing rule, and compare the result with the baseline.
See Rostering working
Book a demo and follow a period from data to forecast to requirement to draft to approval.