06 - Signature Design Decisions

The decisions that defined the product — and why each one was made.

1 — Operations-First Dashboard, Not Chat-First

AVIZ originally started as a pure communication platform — a unified replacement for the WhatsApp and email fragmentation. This was the obvious starting point. But before touching Figma, user interviews produced a finding that reframed the entire product direction.

Operators and brokers who use Leon and FL3XX told me clearly: Their primary daily activity is understanding which aircraft are scheduled for which trips, on which dates, with which passengers — and assigning those trips to the right people. Chat is secondary. It supports the operational picture, it doesn't replace it.

2 — The Three-Pane Split Screen

I started with the hardest UX problem first: how do you give an operator simultaneous access to a trip list, its email thread, and its live broker chat — without overwhelming users who are 30–60 years old and deeply familiar with more traditional interfaces?

The client's initial direction was separate pages for Chat and Email, navigated via a top nav. My position: navigating between pages in a high-pressure operational context isn't just friction — it's a structural failure. Every page switch breaks the user's mental model of the trip they're managing.

3 — The Hidden Task Management Layer

During interviews, the client explained that operator managers frequently assign trips and communication responsibilities to individual team members. At first, this sounded like a simple permissions feature. As I studied how it actually worked in practice, I realised this was an entirely implicit task management process running inside unstructured WhatsApp threads — with no tracking, no confirmation, and no visibility for managers.

I designed a formal task assignment layer on top of the chat system — visible only to operator managers — where broker conversations become assignable tasks with explicit states. The layer is invisible to Sales users, who simply see conversations assigned to them.

Why the workload count in the assignment dropdown matters: In the original WhatsApp workflow, managers had no idea how many active conversations each operator was handling. They assigned based on availability assumptions — which caused some operators to be overwhelmed while others sat idle. Showing task count per person costs the manager zero extra steps and gives them real capacity data for every assignment decision.

4 — The Tag System: Inline vs. Separate Module

Private aviation professionals use a common tag vocabulary: PAX (passenger information), PET (trips with animals), Catering, Special Request. The client's proposal was a dedicated Tags Module — a separate section where users would manually copy messages into the module and assign tags to them. The intent was reasonable. The execution would have been a usability failure.

07 - Operator Panel

Operators think in aircraft. The dashboard was built around that.

This was one of the most deliberate structural decisions in the product. A standard date-sorted list of trips would have forced operators to mentally reconstruct which aircraft was doing what. Instead, the dashboard is organised as a Tail × Date grid — each row is an aircraft (identified by tail number), each column is a calendar day.

An operator scans down their fleet and immediately sees availability, active trips, and assignment status — without navigating anywhere or opening any trip. The recent conversations panel on the right stays visible at all times.

01

Tail × Date Grid Structure

Each row is one aircraft (D-TAIL 1 / 2 / 3). Each column is one calendar day. A single glance tells an operator which aircraft is flying, idle, or unassigned — without opening any record.

02

"Not Assigned" as an Active State

Unassigned cells are explicitly labelled — not left blank. This passive prompt communicates planning gaps to managers without requiring a separate notification or report.

03

Recent Conversations Panel

Always visible alongside the grid. Assignment status shown inline: "You are working", "VBK is working", "HEJ is working" — the whole team's workload visible at a glance, no separate report needed.

Why the grid mirrors Leon and FL3XX: The Tail × Date structure is identical to how operators already think about fleet scheduling in existing aviation tools. This was a deliberate choice to reduce cognitive load — the visual grammar was familiar, only the interactions were improved.

08 - Broker Panel

Brokers manage portfolios, not fleets. The interface reflects that reality.

The broker panel was designed from a fundamentally different starting point. Because brokers typically work individually and don't coordinate within structured team hierarchies, the interface needed to be communication-first — rich in trip context, without the task assignment overhead that would add friction to a workflow that didn't need it.

01

My Trips / Org Trips Toggle

Brokers switch between personal assignments and the full organisational view. Managers use Org Trips to monitor workload distribution across their team — same design pattern, different information density per role.

02

Recent Conversations as First-Class Element

Unlike the operator panel where conversations are secondary to fleet operations, the broker's right panel is the primary communication hub. Their job is communication — the design reflects that priority directly.

03

10-Tab Trip Modal

Trip Details, PAX Data, PET, Catering, Special Request, Slots, Permits, Crew, Invoice, Chatbox. Tabbed structure made it possible to expand scope iteratively as stakeholder interviews revealed additional data categories — without increasing visual complexity.

04

Multi-Leg Trip Structure

Complex charter trips with multiple legs display each leg side-by-side inside the same modal — crew, PAX, EFT, and handling per leg. The broker sees the complete journey at once without additional navigation.

Why the grid mirrors Leon and FL3XX: The Tail × Date structure is identical to how operators already think about fleet scheduling in existing aviation tools. This was a deliberate choice to reduce cognitive load — the visual grammar was familiar, only the interactions were improved.

09 - Design Decisions

Every decision was made for a reason.

Below are the six decisions that most directly shaped the product's architecture and user experience — each framed as the problem it was responding to, the reasoning behind the decision, and the outcome it produced.

10 - Impact & Reflection

What changed — and what it cost to get there.

11 - Key Learnings & Reflections

11 - Key Learnings & Reflections

What this project actually taught me

A 14-week solo design engagement on a first-of-its-kind product produces a specific kind of learning — not theory, but scar tissue. Below are the four things I understood differently by the end of AVIZ than I did at the start.

User interviews shaped the product before design began

User interviews shaped the product before design began

The direction of AVIZ came from user interviews, not from screens or assumptions.
Instead of starting with design tools, I spent time understanding how users actually work and think.

If I had jumped straight into Figma, I would have built a system that looked logical but didn’t match real user behavior. The most impactful decisions in AVIZ came from conversations, not interfaces.

Designing for expert users requires a different approach

Designing for expert users requires a different approach

AVIZ users are professionals who already have strong workflows and mental models.
They don’t need simplification — they need clarity and efficiency.

The goal was not to redesign their workflow completely, but to improve what already works.
Designing something “familiar but better” proved more effective than creating something entirely new.

Saying “no” can be the right design decision

Saying “no” can be the right design decision

One of the key decisions in AVIZ was not implementing a requested feature — inline right-click tagging.

Although requested, it introduced friction and complexity that could negatively impact usability and adoption. Instead, a more structured tagging approach was retained.

This decision was backed by reasoning, usability considerations, and alternative solutions — not just opinion.

Constraints (like security) can improve the product

Constraints (like security) can improve the product

Working within GDPR and banking-level compliance initially felt restrictive.
But these constraints helped shape a more reliable and trustworthy product.

They forced better clarity in areas like data handling, permissions, and user communication — which ultimately improved the overall experience.

Have a project idea in mind? Let’s chat about how we can bring it to life— virtually, from anywhere in the world!