Phenx
Use case · Controlled estimation

Move complex estimates from intake to reviewed proposal.

Organize inspections, documents, operating rules, AI recommendations, and specialist judgment into one visible workflow — without removing estimator or manager authority.

Control boundary

Trigger
New or revised work order
Routine work
Qualification, data assembly, routing
Expert judgment
Scope, specification, and pricing
Consequential control
Manager approves, returns, or stops

Anonymized implementation example

Controlled estimation, in motion

From field evidence to a quote-ready estimate.

See how measurements and field notes become a structured questionnaire; how AI connects recorded conditions to the right product system, technical specifications, and work process; and where experts review uncertainty before a quote moves forward.

Phenx controlled estimation explainer01:02 · Narrated · Captions
The operating path

Every estimate has a visible state, owner, and next decision.

The routine path moves forward automatically. Missing information, hard-stop conditions, and disputed judgment enter a defined exception path.

01

Assemble

Bring measurements, field notes, condition evidence, documents, location, and work-order context together.

02

Structure and qualify

Use AI to turn field evidence into a guided questionnaire, identify missing answers, and apply hard stops before estimating begins.

03

Recommend

Use the structured evidence to recommend the right product system, technical specifications, and work process — and expose uncertainty.

04

Estimate and review

Estimators review scope, assumptions, and price. Managers approve it, return it with a reason, or stop it.

05

Assemble the quote

Produce a quote-ready package while preserving the version, decision history, ownership, and final release authority.

Division of labor

Automation does not need to own every decision.

Rules

Handle certainty

Required information, eligibility, routing, permissions, and non-bypassable stops.

AI

Support judgment

Interpret evidence, classify the work, recommend likely answers, and surface uncertainty.

People

Own consequences

Estimators correct recommendations. Managers retain approval, rejection, and reassignment authority.

Evidence

Verify the path

Record stages, owners, reasons, versions, timestamps, and the correction history.

Exception handling

The exception path is part of the design.

  • Safety and policy conditions can force mandatory review.
  • If no routing rule applies, the work remains pending for a person.
  • Manager rejection returns the estimate through a defined correction cycle.
  • A revised work order creates a new version without erasing prior decisions.
  • If AI is unavailable, the section remains visible and eligible for retry.
Operating improvement

Make expert time more deliberate.

The goal is not to remove judgment. It is to stop consuming expert attention on assembly, routing, and routine decisions that can be controlled another way.

01

Less manual assembly

Bring the relevant information into one operating context.

02

More consistent routing

Use the same qualification and escalation logic across the workflow.

03

Focused expert attention

Direct specialists toward ambiguity, correction, and consequential choices.

04

Visible accountability

See where work is, who owns it, and why it moved or stopped.

Bring your workflow

Which estimate still depends on a few people holding the process together?

In the fit call, we’ll discuss the information, decisions, exceptions, approvals, and operating outcome. No documents or production access are required.