CLOSE
Back To Top

Overview

Aurora is a concept exploring how AI and human specialists could work together to create a premium lifestyle concierge experience. Rather than designing another travel booking app, I explored what it would feel like to delegate an entire experience—from intent and planning to execution and live support—to an intelligent personal assistant.

The project focuses on designing trust. Instead of exposing complex AI processes, the experience makes progress visible through specialised agents that search, verify availability, coordinate logistics and curate personalised recommendations before presenting a seamless plan for approval.

Clickable Prototype


AI CONCIERGE PLATFORM
PRODUCT CONCEPT

Overview

An AI-powered concierge experience that brings intelligent automation and human specialists together to plan and manage highly personalised experiences.

ROLE
Founding Product Designer
SCOPE
Product strategy · UX · UI · AI interaction design · Prototyping
PLATFORM
Mobile customer experience · Specialist desktop workspace
Aurora mobile concierge home screen
Aurora Experience Specialist workspace

Luxury concierge is inherently high-touch. Customers expect the service to understand their preferences, anticipate what they need and coordinate multiple people and services without making them manage the complexity themselves.

The challenge was not simply:

“How might AI help someone plan a trip?”

It was:

“How might AI and human specialists work together behind the scenes
while giving the customer one simple, trusted experience?”

Delegate, don’t browse

Rather than asking customers to search through endless options, Aurora starts with intent and handles the complexity behind the scenes.

Human judgement, amplified

AI recommends, coordinates and surfaces context while specialists exercise judgement where expertise matters.

Bring customers in at meaningful moments

Progress happens quietly in the background. Attention is requested only when a genuine decision is needed.

Simplicity for the customer.
Context and control for specialists.

I wasn’t designing two independent products. I was designing two views into the same service.

01

Customer experience

The customer expresses intent, receives reassurance and is brought back in only when a decision is needed.

Aurora customer chat experience

Intent

Aurora Intelligence

Specialist

02

Specialist workspace

The specialist sees the context, constraints, recommendations and coordination needed to make a high-quality decision.

Aurora Experience Specialist dashboard

From intent to execution.

01

Home

An ongoing concierge relationship.

Aurora home screen
02

Voice / conversation

Express what you want naturally.

Aurora voice input screen
03

Context & direction

Aurora understands the occasion, dates, preferences and relationship context.

Aurora conversational planning flow
04

Specialist orchestration

AI and specialists work together behind the scenes.

Experience Specialist workspace
05

Proposal

A coherent experience emerges from the orchestration.

Aurora Paris anniversary proposal
06

Approval

One meaningful decision instead of dozens of micro-decisions.

Aurora trip approval screen
07

Confirmation

Aurora handles execution.

Aurora trip confirmed notification
08

Itinerary

Practical detail appears when it becomes useful.

Aurora Hidden Paris itinerary

From a request
to an understood intent.

Aurora doesn’t treat the customer’s message as an isolated prompt.
It recognises the occasion, dates and preferences, considers Catherine’s profile and then translates that context into three distinct experience directions.

Instead of returning an endless list of hotels and restaurants, the system helps the customer make a higher-level decision first.

Use AI to reduce decisions, not create more of them.
Aurora understanding customer intent

The system behind the simplicity.

Behind the customer experience, specialists need enough context to make
high-quality decisions without repeatedly asking the customer for information
Aurora already knows.

Aurora Experience Specialist system

01

Understanding the moment

Occasion, pace, preferences and relationship context are surfaced before curation begins.

02

Explainable recommendations

Aurora Intelligence shows match scores and why an experience fits the customer.

03

Human judgement remains visible

Specialists verify availability, assess alternatives and confirm the final selection.

AI recommends.
Humans exercise judgement.

One meaningful approval instead of dozens of micro-decisions.

Aurora works in the background until everything is ready. The customer is then brought back in with a simple notification and a consolidated proposal covering flights, stay, dining, experiences and logistics.

Pricing remains transparent, but the experience avoids feeling like a traditional travel checkout.

Trip ready for approval notification
Aurora trip approval screen
Aurora trip confirmed notification

The complete product language.

DESIGN SYSTEM

A system designed
to feel distinctly Aurora.

I created a lightweight design system to give Aurora a consistent
product language across the customer and specialist experiences —
balancing the warmth and editorial character of luxury hospitality
with the clarity required of an AI-powered product.

Aurora design system colour palette

01

Colour

A warm gold accent creates a sense of understated luxury against deep, near-black surfaces. Supporting colours are used intentionally for status, progress and system feedback.

Aurora design system typography

02

Typography

Editorial serif typography brings warmth and personality to moments of discovery, while a clean sans-serif supports functional UI, conversation and complex information.

Aurora design system components and corner radii

03

Components

Shared radii, surfaces and reusable components create consistency across mobile and specialist interfaces, while allowing each experience to respond to its different context and information density.

Not a chatbot.
Not a travel app.
A service orchestration layer.

The interesting design problem wasn’t how Aurora should talk. It was how the product should understand intent, retain context, coordinate people and systems, communicate progress, surface decisions and ultimately make a complex premium service feel effortless.

Delegate, don’t browse

Customers express intent rather than manually assembling solutions.

Human judgement remains visible

AI supports specialists rather than pretending expertise can always be automated.

Bring customers in at meaningful moments

Attention is requested only when a genuine decision is needed.

Figma

FigJam

AI-assisted ideation

Interactive prototyping