EPIPHANY AI3 LLC / TUCSON, ARIZONA

Intelligence you can open, inspect, and service.

AMANDA is a smart-kitchen concept organized around a visible utility spine: modular appliances, water systems, and service access designed as one maintainable environment.

CONCEPT ANIMATION / 00:20V01 — 2026

Concept visualization. Features and performance remain subject to engineering validation.

THE OPERATING IDEA

Built around the service call—not hidden from it.

01

Maintainable

Service access is part of the architecture.

02

Measurable

Every performance claim needs a test method.

03

Modular

Hot, utility, and cold systems can evolve independently.

CONCEPT ARCHITECTURE / FOUR SYSTEMS

Inside the utility spine.

Each element below is a design hypothesis shown in the animation—not a statement of built performance.

HOT SIDETHERMAL
AMANDA COREUTILITY
COLD SIDECOOLING
A

SERVICE ACCESS

360° service carousel

A rotating lower module organized around a stationary utility point, conceived for direct inspection and replacement access.

B

WATER STEWARDSHIP

Water management loop

A visible separation, filtration, heating, and drain-path concept that treats water as a managed system rather than hidden plumbing.

C

ADAPTIVE WORKSPACE

Concealed appliance lifts

Countertop appliances move between a protected parked position and an accessible service or show position.

D

MODULAR SYSTEM

Zoned utility spine

Hot-side, utility-core, and cold-side systems are separated into legible modules for testing, maintenance, and future iteration.

EVIDENCE GATE / CLAIMS CONTROL

No claim crosses the line without proof.

AMANDA separates the promise of a concept from the evidence needed to market a product. Accuracy, efficacy, savings, safety, and sustainability claims stay provisional until the supporting record is competent, reliable, and retained.

01

Define

Write the exact claim, intended user, conditions, and known limitations.

02

Test

Use a repeatable protocol with a baseline, failure cases, and acceptance criteria.

03

Retain

Preserve results, configurations, dates, exceptions, and decision records.

04

Disclose

Publish only what the evidence supports—along with boundaries and uncertainty.

AI OPERATING BOUNDARY

Useful intelligence.
Defined authority.

An AI feature may assist, alert, explain, or recommend. It does not receive unlimited authority over people, safety systems, or the truth.

01

SAFE + LAWFUL

No dangerous or illegal use, rights violations, or unconsented tracking.

02

SECURE BY DESIGN

No system behavior that enables abuse, disruption, malware, or bypass of safeguards.

03

HUMAN AUTHORITY

No materially harmful high-impact decision without accountable human supervision.

04

TRUTHFUL PROVENANCE

No deceptive capability, expertise, identity, or human-authorship claims.

POLICY REFERENCE

Boundaries adapted from the supplied Google Generative AI Prohibited Use Policy snapshot, last modified December 17, 2024. This is a design reference—not a statement of Google affiliation or endorsement.

View official policy

INVENTION RECORD / DECISION PATH

Concept to protectable, transferable value.

The animation communicates a system. The next job is to separate visual story, technical function, brand, and confidential know-how—then build the right record for each.

01

Capture

Record the problem, architecture, alternatives, dates, and each contributor’s role.

02

Classify

Separate potentially patentable function from brand, creative expression, and confidential know-how.

03

Search

Investigate prior art and competing systems before narrowing what may be new.

04

Decide

Choose what to file, publish, hold as a trade secret, license, test, or retire.

05

Prepare

Build an attorney-, agent-, or pro-se-ready evidence packet before any filing decision.

IP01

USPTO / PATENTS TOOLKIT / VOLUME 1

Start with the foundations.

The supplied USPTO toolkit introduces patent basics, IP categories, inventorship, utility-patent qualification, account setup, and filing terminology for independent inventors and small businesses.

Open the toolkit

SOURCE LIBRARY / PROVIDED RECORDS

Inspect the inputs.