Amulet AI

AI advisory and implementation for Australian businesses

AI for the work that refuses to stay simple.

Amulet maps the systems, source material, exceptions and human decisions behind difficult work, then builds practical AI around the operating path.

Bring us a hard problem Inspect the capability

30 minutes · problem-fit request · no software brief required

Position

The tool is not the starting point. The business is.

Amulet helps Australian businesses fix work slowed by manual steps, disconnected systems or hard-to-use information. We use the software already in place, build what is missing, and keep people in charge of approvals and commitments.

Proof of method

See the standard of thinking before you engage us.

These reference artefacts show how Amulet frames the work. They are inspectable illustrations of the method, never client evidence.

Opportunity review excerpt

A decision question, not a technology shopping list.

Observed signal
Variation work crosses email, drawings and finance records before a commercial decision can be made.
Decision question
Can the source documents and assumptions be reconciled before the commercial owner reviews the change?
Evidence expected
A source map, exception list, decision owner and bounded first-step options.
Reference workflow architecture

The exception and approval path are part of the design.

Systems + data
Inbox, document store and finance record flow through ingest, extraction and reconciliation.
Evaluation + exceptions
Completeness, source version and business rules are checked; conflicting evidence stops the path.
Human approval
The named commercial owner decides before any external commitment.
Implementation decision record

Scope, assumptions and recovery remain inspectable.

Scope
Prepare a draft variation pack from approved source records.
Assumptions
The referenced source version and rate table are current.
Recovery + evidence
Stop safely, retain inputs, record the reason, approver and authorised output.
Source records follow a controlled route through review and retained evidence.

Client context

Configured around how the business works.

Discovery maps the workflow before implementation: what information matters, where decisions sit, which systems are authoritative and what must remain human-controlled.

Find the right work
Understand where friction sits and what is worth doing.
Build what is missing
Design practical capability around approved systems and people.
Make it useful
Launch with the right reviewers, measure the outcome and improve.

Capability

From a business problem to working AI.

Amulet combines operational context, software engineering and implementation at the scope the business needs.

  1. 1

    Discover

    Understand the work and what is worth doing.

  2. 2

    Design

    Agree the workflow, controls and named decisions.

  3. 3

    Prepare

    Structure approved documents, data and images.

  4. 4

    Connect

    Integrate systems and select the right AI model.

  5. 5

    Support

    Present clear outputs, sources and exceptions.

  6. 6

    Embed

    Pilot with users, measure and extend where useful.

Explore the complete capability

Product direction and custom delivery

Different business needs. One Amulet approach.

Every client needs business-context configuration. Amulet’s planned product families and current custom workflows are being shaped around the same configurable approach, not separate software forks.

See the product direction

Controlled delivery

Amulet prepares.
The named person decides.

Consequential actions follow the boundary agreed for the workflow. The approver sees the sources, prepared action and resulting record together.

  1. 1
    Approved source context

    The current system remains authoritative where it works.

  2. 2
    Prepared work

    Context, assumptions and missing information remain visible.

  3. 3
    Named human decision

    A person approves consequential action.

  4. 4
    Activity & evidence

    The source, approver and authorised outcome stay connected.

Engagement path

Commit in stages. Increase investment when the evidence earns it.

The first conversation validates fit. The AI Business BEEP creates a decision basis. Implementation remains a separate go or no-go commitment.

Inspect the engagement path

Start with the problem

Find the first piece of work worth fixing.

Bring the workflow, the friction and the decision that matters. We’ll start there.

Bring us a hard problem