← Back to work 04 / 04
Data Visualisation SaaS Microsoft Platform Enterprise

KnowledgeKube

UX design for a Microsoft-based self-service dashboard platform — transforming enterprise Excel data into compelling, industry-specific visualisations for clients across holidays, transport, infrastructure, and beyond.

Mercato — KnowledgeKube
NDA Protected

Client dashboards and proprietary platform designs cannot be shared under the terms of this engagement. The process and contributions documented here reflect my actual work on this project.

5+ Distinct industry verticals — holidays, transport, software, infrastructure, architecture
0 Lines of code required from clients — Excel upload to live dashboard
Microsoft Platform foundation — tight integration with Microsoft's ecosystem and data tooling

Overview

KnowledgeKube (developed by Makato) is a Microsoft-based SaaS platform that allows enterprise clients to transform their spreadsheet data into rich, interactive dashboards — without any technical knowledge or development resource. The core proposition is simple: upload your Excel file, get a beautiful, branded dashboard that makes your business data immediately legible and actionable.

I joined the design team as a UX Designer, working alongside other designers on a client portfolio that spanned high-profile organisations across multiple industries. My focus was on translating each client's specific data structures and business objectives into dashboard designs that communicated the right insights at a glance.

01 Discover

Understanding what businesses actually need from their data

The challenge with data visualisation is not making data look good — it's understanding which story each dataset needs to tell, and who in the business needs to read it.

Client Discovery Sessions

Each new dashboard engagement began with a structured discovery session to understand the client's business context, key metrics, decision-making hierarchy, and how the data was currently being used — usually a combination of manual spreadsheet analysis and verbal reporting. Identifying the gap between what the data contained and what decisions it needed to support was the starting point for every design.

Data Audit

Reviewed client Excel files to understand data structure, completeness, and the quality of existing labelling. Data that seems simple in a spreadsheet often reveals complexity when examined for visualisation: inconsistent categorisation, mixed date formats, and unlabelled columns were common. Understanding the data at this level shaped both the design and the data preparation guidance we gave back to clients.

User Role Mapping

Mapped who within each client organisation would be viewing the dashboard and for what purpose — C-suite overview vs. operational day-to-day monitoring vs. frontline team performance. Different roles required different levels of granularity and different visualisation types, and this mapping drove decisions about hierarchy and drill-down depth.

Industry Benchmarking

Researched how leading platforms in each client's sector presented comparable data — what chart types were conventional, what KPIs were standard, and where there was opportunity to do something more distinctive. For clients in industries like architecture or infrastructure, the visual language of the dashboard needed to feel at home in boardroom presentations.

02 Define

From raw data to a clear visual brief

Before a single chart was drawn, I needed to define what each dashboard needed to communicate — and what it didn't.

Priority metrics framework

For each client engagement, I produced a structured brief that ranked the metrics by decision-making importance — separating the key figures that needed to be immediately visible from supporting context that could sit a level deeper. This prevented the common failure mode of dashboard design: trying to show everything at once and communicating nothing clearly.

Visual language and brand alignment

KnowledgeKube dashboards needed to feel premium and credible for enterprise clients — many of whom would be presenting them internally to senior leadership. I worked within each client's brand guidelines to establish colour systems for data categories, typography hierarchy, and the overall visual weight of the dashboard. The goal was a result that felt deliberately designed, not automatically generated.

Interaction model for non-technical users

The platform's self-service promise meant that clients — not developers — would be maintaining and updating dashboards after handoff. This shaped every design decision: interaction patterns needed to be learnable without training, and the relationship between the source Excel data and the displayed visualisation needed to be transparent enough for a non-technical user to understand what would happen when they updated a spreadsheet cell.

03 Design

Designing across industries — the same challenge, different answers

Every client brought a different data story. The skill was in knowing which visual approach served each story best — and resisting the temptation to apply the same solution twice.

Hospitality and holidays — performance at a glance

Holiday and hospitality clients needed dashboards that surfaced booking performance, occupancy rates, and revenue trends in a format their operations teams could action daily. The design challenge was making time-series data scannable across multiple locations simultaneously — using small-multiple chart patterns and RAG (red-amber-green) status indicators to communicate relative performance without requiring detailed reading.

Transport and infrastructure — precision and compliance

Transport and infrastructure clients operated in regulated environments where data accuracy and audit trails mattered as much as visual clarity. Dashboard designs for these sectors prioritised precision — showing exact figures alongside visual summaries, with clear data source and refresh-time indicators. These clients also typically had longer reporting cycles, so the dashboard design needed to accommodate both real-time monitoring views and period-end summary layouts from the same underlying dataset.

Architecture and professional services — presentation-grade design

Architects and professional services firms used their KnowledgeKube dashboards in client-facing contexts — presenting project status, resource allocation, and financial performance in client meetings. This required a higher standard of visual polish and demanded that the dashboard feel like a designed artefact rather than a software tool. I developed board-ready layouts with print and presentation export in mind, using white space and typographic hierarchy to create reports that could stand alone without verbal explanation.

Software and technology — developer-friendly yet executive-ready

Technology sector clients often had the most data-literate users but the widest range of stakeholders — from engineers who wanted granular detail to executives who needed high-level KPIs. I designed dual-mode dashboard layouts that could toggle between a summary executive view and a detailed operational view from the same page, using progressive disclosure to serve both audiences without requiring separate builds.

04 Deliver

What I took from data visualisation at scale

Working across a diverse client portfolio in a design team forced a discipline that's easy to neglect in single-product environments: the ability to shift mental model quickly. An architecture firm and a transport operator have fundamentally different relationships with their data, different visual cultures, and different definitions of what "useful" looks like on a screen.

The biggest lesson from this engagement was that data visualisation is primarily a communication problem, not a design one. The most elegant chart is worthless if it's answering a question nobody is asking. Getting to the right question — through discovery and stakeholder alignment — is where most of the value is created.

KnowledgeKube's core promise — upload your Excel, get a dashboard — put enormous UX pressure on the design to feel effortless. The complexity of data processing and visualisation logic happened entirely behind the interface. My job was to make sure nothing of that complexity leaked through to the user.

Back to Featured

Airbnb True-to-Listing

Trust Platform · AI Integration · 2024