Phenx
Use case · Controlled purchasing

Turn fragmented inventory into approved, traceable transactions.

Combine operating rules, AI prioritization, human financial approval, and RPA execution to move through high-volume purchasing decisions without surrendering control over what gets bought or at what price.

Control boundary

Trigger
New inventory and buying demand
Routine work
Collection, filtering, reconciliation
Expert judgment
Candidate selection and bid amount
Consequential control
A person confirms total exposure

Anonymized implementation example

Controlled purchasing, in motion

From fragmented evidence to an approved transaction.

See how operating rules narrow the field, AI prioritizes the candidates worth attention, a person approves quantity and financial exposure, and RPA executes only the approved path — stopping when the item, page, or session does not match.

Phenx controlled purchasing explainer00:57 · Narrated · Captions
The operating path

Move from inventory signal to approved action.

Separate repetitive collection and execution from the financial judgment that should remain visible, reviewable, and owned by a person.

01

Collect

Assemble inventory, valuation, condition evidence, reports, images, and purchase history.

02

Narrow

Apply buying plans, required conditions, location, age, price, and other operating constraints.

03

Prioritize

Use AI to score candidates and direct operator attention toward the strongest possibilities.

04

Approve

Let operators inspect evidence, edit prices, remove items, and confirm quantity and total exposure.

05

Execute and verify

Use RPA to perform approved transactions, then record status, screenshots, and purchase history.

Division of labor

Automate execution without automating financial authority.

Rules

Enforce the buying boundary

Apply required conditions, inventory constraints, buying plans, and hard filters consistently.

AI

Prioritize attention

Rank candidates using available operating signals without making the final commitment.

People

Approve the exposure

Inspect evidence, override amounts, remove candidates, and confirm the basket before action.

RPA

Execute the approved action

Navigate the transaction system, enter approved values, submit, and return execution status.

Evidence closes the loop: bid histories, execution statuses, screenshots, and purchase records show what happened.

Exception handling

Stop uncertainty before it becomes a transaction.

  • If an item or transaction page does not match, execution stops with an error status.
  • Invalid sessions are detected, reset, and retried rather than silently ignored.
  • Operators can edit prices, remove candidates, or clear the basket before approval.
  • A confirmation step exposes transaction count and total financial value.
  • Execution records and screenshots make failed or disputed actions reviewable.
Operating improvement

Let people decide. Let automation do the repetition.

The goal is a faster, more consistent buying operation in which financial authority remains explicit and execution evidence remains available.

01

Less manual searching

Aggregate and narrow fragmented inventory before an operator begins reviewing.

02

Consistent buying criteria

Apply the same operating constraints before candidates reach the decision point.

03

Explicit financial control

Keep candidate selection, price changes, and total exposure visible before execution.

04

Traceable execution

Connect the approved intent with status, evidence, and the resulting purchase record.

Bring your workflow

Which transaction still depends on people searching, copying, and clicking?

In the fit call, we’ll discuss the information, buying rules, approval boundary, execution system, exceptions, and financial consequence. No production access is required.