Demo

From risky input to decision evidence in one controlled flow

See how PalmerAI handles prompts and documents through policy, approval, and audit-ready evidence. The demo shows the governed path from intake to decision record without hiding behind generic product language.

Truth-first demo walkthrough
One governed path across prompts and documents
From uncontrolled usage to governed evidence

This walkthrough stays product-realistic: uncontrolled start, governed intake, document checks, approval, tenant context, decision record, evidence, and stable oversight.

Segment 1: The gap Uncontrolled requests and file handling bypass structure until they enter one governed intake path.
Segment 2: The control layer Inspectability, policy fit, approval state, and tenant context remain attached during handling.
Segment 3: The evidence path Decision state, evidence fields, and stable oversight remain reviewable later for buyers and operators.

This walkthrough uses approved PalmerAI source frames only. No generated UI or concept-art dashboard is involved.

Main demonstration asset

This is the primary stitched product demonstration for the buyer path. It shows one governed sequence rather than a documentation gallery.

Exact runtime states

Request intake, document checks, approval required, decision record, evidence ready, and stable oversight stay visible in sequence.

Truth-first only

Every screen in this demo comes from approved PalmerAI source frames. Logos, labels, and UI surfaces are fixed from source assets only.

Proof frames

Static source frames under the stitched motion

These chapter frames stay below the main demo asset as proof of the exact source visuals used in the motion sequence.

PalmerAI request received frame showing governed request intake.

Request received

Requests stop bypassing the control layer and enter a named governed path before model-side execution begins.

PalmerAI document checked frame showing inspectability and class signals.

Document checked

Inspectability, type, classes, and policy fit remain visible before the workflow can continue.

PalmerAI approval required frame showing explicit approval state.

Approval required

High-risk or uninspectable inputs create an explicit approval object instead of informal side-channel review.

PalmerAI decision recorded frame showing reviewable evidence fields.

Decision recorded

Decision state, reason codes, and timestamps stay legible later instead of getting rebuilt from raw logs.

PalmerAI evidence ready frame showing exportable review state.

Evidence review-ready

Evidence exports stay clean and reviewable with hashes, timestamps, policy version, and approval state attached.

Flow Walkthrough

1. Request enters

A request or document enters the gateway instead of being sent straight to a model endpoint.

2. Document checked

PalmerAI checks policy, document type, inspectability, and class signals before anything moves forward.

3. Approval required

Risky or uninspectable events create explicit approval objects rather than informal human workarounds.

4. Decision logged

The final state records outcome, actor, timestamps, and policy version for later review.

5. Evidence exported

The final state includes decision outcome, reason codes, timestamps, policy version, and supporting metadata.

What Each Role Sees

Operator

Reviews approvals, incidents, document evidence, and cross-tenant governance posture.

Customer admin

Handles use-case-level approvals and sees what entered the workflow under their tenant.

Partner

Sees tenant-level governance summaries without exposing raw document contents in routine reporting.

Use the demo to see how PalmerAI stays controlled in live workflows.

What matters is the same governed path holding across request intake, document control, approval, and evidence export. Use the evidence pack and one-pager when buyers need the audit and commercial context quickly.