Product Design Brief · Enterprise AI Workflow

Clarendon
Asset Management Platform

AI-assisted asset management, role-based workflows, and audit-ready product design

Project
Clarendon Federal Asset Platform
Deliverables
Personas, interaction flow + live platform prototype
Document Type
Product Design Brief / Design Rationale
Year
2025
Interactive · Skip to Prototype Try the live Clarendon experience →
01 · Key Personas
02 · Interaction Flow
03 · Interactive Prototype
01
Deliverable · Key Personas

Five Roles,
One Accountable System

Five federal user types — each with different goals, risk tolerance, and interaction patterns — informed the platform structure, AI transparency model, and audit-ready workflow design.

Dana Brooks · Operations Manager
Marcus Hill · IT Director
Elena Ruiz · Facilities Coordinator
Priya Shah · Auditor / Compliance
Thomas Green · Department Admin
Clarendon Key Personas

Loads from horizonsplash.com

Deliverable 01 · Key Personas
Role-Based Needs,
Shared System Trust

Each persona was defined by operational responsibility, success criteria, trust barriers, and decision rights — surfacing design requirements before layout, navigation, or AI interaction patterns were explored.

↗ Open Full Size
View Deliverable
clarendonpersonas.png
Design Intent

One Platform, Five Realities

The same interface serves an auditor validating compliance records and a facilities coordinator tracking equipment location. Each persona surfaces a different layer of the system's complexity.

Key Tension

Trust vs. Speed

Priya needs audit-ready traceability; Dana needs fast answers. Designing for both meant AI outputs needed clear sourcing, not just results — so speed didn't come at the cost of accountability.

Key Finding

AI Skepticism Is a Design Problem

Marcus distrusts AI results and Elena has low digital comfort. Black-box answers aren't acceptable. Every AI response in the interface links back to its source record.

02
Deliverable · Interaction Flow

From Dashboard Scan
to Audit Trail

The core workflow across all five personas — from first login through search, asset detail, AI query, and audit-ready activity log. The design supports fast scanning without sacrificing traceability.

01
🚀
Onboarding
02
📊
Dashboard
03
🔍
Filter & Search
04
🗂️
Asset Registry
05
📋
Detail View
06
🤖
AI Query
07
📈
Activity Log
Clarendon Interaction Flow

Loads from horizonsplash.com

Deliverable 02 · Interaction Flow
Nonlinear Use,
Clear System Logic

Seven steps designed so each persona can enter the platform at the moment most relevant to their role and still leave with a traceable, defensible result. The workflow is linear enough to understand and flexible enough for enterprise reality.

↗ Open Full Size
View Deliverable
clarendonflow.png
Design Intent

Role-Based Entry Points

Not every user starts at Step 1. Thomas sets up permissions and leaves. Priya jumps straight to the Activity Log. The flow is linear by default but non-linear by design.

Key Decision

AI Query Is Step 6, Not Step 1

The AI query sits after users have already found and opened an asset — so natural language questions have context, not just keywords. It's a refinement tool, not a search replacement.

Key Finding

Activity Log Closes the Loop

For auditors and managers, Step 7 is the most important screen. The flow ends with traceability — who did what, when, and why — which is the trust signal the compliance persona needs most.

03
Deliverable · Interactive Prototype

Clarendon Platform
Live Product Prototype

Live desktop prototype showing the core platform experience: dashboard scanning, filtering, asset registry, record detail, AI query, and activity log. The prototype demonstrates how product structure supports speed, trust, and accountability.

horizonsplash.com/protoarchives/clarendon/clarendon.html
Deliverable 03 · Interactive Prototype
Enterprise Workflow
Prototype

Navigate the full 7-step flow directly in the desktop frame. Interact with filters, the asset registry, and the AI query interface as they were designed to function in production.

↗ Open Full Screen
Open Prototype
horizonsplash.com · clarendon.html
What to Look For

Dashboard Scanning

Notice how KPIs and alerts are surfaced before any action is required. Dana and Marcus should be able to identify a problem within seconds of logging in — no drilling required.

What to Look For

AI Query Transparency

When running a natural language query, observe how the response links back to source records. This is the trust mechanism designed to address Marcus's skepticism and Priya's compliance needs.

What to Look For

Audit Trail Completeness

The Activity Log at the end of the flow should feel conclusive — a complete record of every action taken during the session, exportable and timestamped for compliance purposes.