Receives
- Accounting and operational exports
- Budgets, forecasts and assumptions
- Volume and workforce files
- Periodic reporting requests
Make data collection more reliable and prepare variance analysis.
The AI employee assembles data, checks repositories, prepares variance commentary and keeps a record of assumptions. The controller retains interpretation and recommendations.
Controlled executionSUBSTI / RUN-284
- 01Mission
Supplier invoice received
- 02Plan
Read · reconcile · enter · verify
- 03Memory
Procedure AP-04 and supplier record
- 04Tool
ERP opened through a controlled session
- 05Exception
Purchase order variance detected
- 06Human
Validation requested above threshold
- 07Trace
Action closed and execution logged
Operating model illustration. This is not a product screenshot.
AI employee scope
From incoming information to a controlled hand-off.
Understands
- Period and scope
- Analytical dimensions
- Allocation rules
- Formats and variance thresholds
Searches
- Reference sources
- Approved assumptions
- Prior-period commentary
- Operational events explaining variances
Executes
- Consolidate authorised files
- Prepare consistency checks
- Calculate variances using approved formulas
- Pre-fill sourced commentary
- Produce expected tables
Checks
- Comparable periods and scopes
- Totals and subtotals
- Repositories and units
- Series breaks
- Presence of a source for each assumption
Hands to people
- Variance without sufficient explanation
- New assumption
- Scope change
- Budget trade-off
- Management communication
You define the autonomy of your AI employees.
Assistant
The AI employee researches, analyses and prepares. A person performs the action.
- Research
- Analyse
- Prepare
Preparer
The AI employee prepares a complete action and waits for approval before execution.
- Prepare the action
- Explain checks
- Wait for approval
Operator
The AI employee performs authorised actions, flags exceptions and preserves a trace.
- Execute authorised actions
- Check the result
- Flag exceptions
Manager
The AI employee coordinates several processes or AI employees and involves people at critical stages.
- Coordinate
- Apply rules
- Manage dependencies
- Escalate critical stages
Exceptions
What stops or escalates.
- Missing file or changed structure
- Misaligned repository
- Late data
- Unapproved formula
- Out-of-threshold variance
KPIs
What to measure.
- Collection time
- Number of repository corrections
- Reporting production time
- Share of sourced variances
- Time spent analysing rather than preparing
Integration
What it can use.
- ERP and accounting
- EPM, BI and reporting tools
- Spreadsheets and document workspaces
- Source operational systems
Compatible processes
Begin with one bounded and repeatable mission.
01Periodic data collection
02Repository checks
03Preliminary variance analysis
04Table consolidation
05Forecast preparation
Indicative subscription: around €1,000 per month per AI employee, plus integration setup. Final scope depends on software, volumes, controls and infrastructure.
FAQ
Questions before scoping.
Is variance commentary generated without review?
It can be prepared from identified facts and sources. Interpretation, new assumptions and management communication are approved by the controller.
Can it work with different Excel files each month?
Representative structures and their variations are tested. An unexpected change can trigger a stop and review rather than a silent consolidation.
Which role should work differently tomorrow?
Tell us about a function, a volume and the software involved. We will identify the first scope to turn into an AI employee.
Book a demo