Estimate your savings

Start with a typical process, then adjust the numbers to match yours. The results update as you go.

Your process

The proportion of tasks an agent could complete, with a person stepping in for the rest.
Time for a person to check or approve the agent's work.
Fully loaded cost in AUD, including on-costs.

Your estimated savings

Hours saved per month
Hours saved per year
Value of time saved per year
Full-time equivalents freed up

Scenarios

Savings under conservative, expected and optimistic scenarios
Scenario (share handled)Hours a yearValue a yearFTE
Assumptions
  • Conservative and optimistic scenarios use 60% and 125% of the share you entered (capped at 95%).
  • One full-time equivalent is 1,672 hours a year (38 hours a week for 44 weeks).
  • Build and running costs aren't included. This is an indicative estimate, not a quote.

Get a free review of your estimate

A consultant will sense-check your numbers, suggest the best process to start with and outline what a pilot would involve.

Ready to test it on a real process?

Frequently Asked Questions

Hours saved per month = tasks per month × share handled by the agent × (minutes per task − review minutes) ÷ 60. Annual value multiplies the yearly hours by your hourly cost. One full-time equivalent is 1,672 hours a year (38 hours a week for 44 weeks).

Good AI agents bring a person in where it matters. Review time keeps the estimate realistic by allowing for someone to check or approve the agent's work.

Use the fully loaded cost of the people doing the work today, including on-costs such as superannuation, leave and overheads, not just their salary.

No. It estimates the time and cost freed up. We'll help you compare that with build and running costs when we review your estimate.

No. The estimate is calculated in your browser. If you ask us to review it, we store your contact details and inputs in our CRM so a consultant can follow up.

Speak to an AI Agents Expert

Have a process in mind? We'll help you pick the best place to start, design an agent with the right guardrails and measure the results.

Contact an AI expert