Case Study · AI Product Design

Krayo — AI-Powered
Finance Platform

Designing Human-AI workflows for enterprise finance teams — simplifying procurement, spend, compliance, and HR for CFOs, finance leaders, and operators.

Role
Lead Product Designer
Duration
6+ Months
Product
Enterprise SaaS · Fintech + AI
Users
CEOs, CFOs, Finance Teams
Clients
6 Enterprise + 1 Unicorn

⚡ TL;DR for recruiters: Designed an AI-powered finance platform used by 6 enterprise clients and a unicorn company. Led end-to-end UX across spending, procurement, compliance, and HR modules — improving financial processing efficiency by 32% and contributing to 18% revenue growth through better workflow design and executive usability.

32%
Improvement in financial processing efficiency
18%
Revenue growth contribution through workflow optimisation
7
Enterprise clients including 1 unicorn company
The Problem

Finance teams were drowning, not lacking data.

Spending, procurement, compliance, and HR systems were completely disconnected. Financial approvals were manual and time-consuming. Executives lacked real-time, actionable insights. And AI adoption was low due to deep trust and usability concerns.

The result: slow decisions, operational inefficiency, and missed financial opportunities — even with data available.

◈
Fragmented Workflows
Spending, procurement, compliance, and HR lived in separate systems with no unified view.
◎
Manual Approvals
Financial approvals required multi-step manual handoffs, slowing decision cycles significantly.
▦
No Executive Clarity
CEOs and CFOs lacked real-time, actionable insights — only raw data dashboards.
⬡
Low AI Trust
AI adoption was blocked by users not understanding or trusting AI-generated recommendations.
My Role

End-to-end ownership — from discovery to production.

I led product design for the entire Krayo finance platform — defining UX strategy, designing across all modules, building the design system, and establishing AI interaction guidelines that became internal standards.

I worked directly with Product and Engineering in agile sprints, conducting enterprise usability testing to validate AI-driven features before production release.

01
End-to-end product design for AI finance platform
02
UX strategy for Spending, Procurement, Compliance & HR
03
Scalable design system for AI-first features
04
AI interaction & prompt design guidelines
05
Enterprise usability testing + agile sprint collaboration
Research & Insights

What I heard from finance leaders.

I conducted enterprise stakeholder interviews with CFOs and finance teams, mapped financial processes end-to-end, ran usability testing on AI features, and iterated based on real feedback before any production release.

◈
"Give us clarity, not more dashboards"
CFOs need quick, high-level financial clarity. Raw data dashboards create cognitive overload.
◎
Approval workflows were the biggest pain
Finance teams flagged multi-step approval workflows as their single biggest source of frustration.
▦
Trust requires transparency + control
Users only trusted AI when they could see why a recommendation was made and edit or override it.
⬡
Guided AI beats open-ended AI
Enterprise users strongly preferred structured, guided AI workflows over chat-style open interfaces.
Design Process

From sketches to production.

The process moved from deep discovery through iterative sketching, system design, and validated production deployment — always with enterprise users in the loop.

01
Discovery
CFO and finance team interviews. Workflow mapping. Pain point analysis.
02
Sketch & Wireframe
Hand-sketched wireframes for procurement policies, spend pipelines, analytics funnels.
03
Design & System
High-fidelity UI across all modules. Built scalable component library and AI guidelines.
04
Test & Ship
Enterprise usability testing with real finance teams. Deployed to 7 clients.
Design Artefacts

From napkin to screen.

Wireframe sketches show the thinking behind the design — procurement policies, workflow analytics, spend pipelines — before they became production UI.

Procurement Policy Wireframe
Wireframe · Procurement

Create New Procurement Policy

Hand-sketched wireframe exploring procurement policy creation — title, effective dates, approver pre-sets, applicability filters, and AND/OR logic strips.

Workflow Analytics Wireframe
Wireframe · Analytics

Procurement Workflow Analytics

Funnel visualisations by number of requests and amount across workflow stages, plus TAT chart showing time taken per stage.

Spend Pipeline Wireframe
Wireframe · Spend Pipeline

Workflow Summary & Spend Pipeline

Active workflow summary with key stats, spend pipeline funnel chart, kanban-style request cards, and filter controls.

Completed Workflows Wireframe
Wireframe · Completed Views

Completed & Closed Workflow Views

Completed workflows broken down by Category, Function, and Quarter with pie chart visualisations.

Software Spend Dashboard
Final UI · Software Spend

Software Spend Dashboard

Executive-facing overview with committed annual spend, renewals calendar, spend progression vs budget, and AI Insights panel.

Procurement Module
Final UI · Procurement

Procurement — Active Requests

Spend pipeline funnel, kanban board with stage counts, colour-coded request cards, and filter controls.

Vendor Master Dashboard
Final UI · Vendor Master

Vendor Master — Summary View

Vendor intelligence with renewal calendar, donut charts, geo distribution map, and top vendor table.

App Detail View
Final UI · App Detail

App Detail & Analytics

Deep-dive app view showing contract value, engagement score, monthly spending chart, and value score trends.

Documents & Agreements
Final UI · Documents

Agreements & Document Management

Centralised document view with AI Insights surfacing duplicate solutions and upcoming commitments.

1 / 4
The Solution

Five design decisions that moved the needle.

1
AI as Co-Pilot, Not Replacement
Embedded AI assistance across all modules — not as an autonomous agent, but as a co-pilot that suggests, explains, and lets users override.
AI WorkflowsTrust Design
2
Executive-First Dashboard Design
Designed for CFOs and CEOs: real-time financial snapshots, predictive AI insights, risk alerts — high-signal, low-noise.
Executive UXData Viz
3
Guided AI Interactions
Pre-defined prompt structures, context-aware suggestions, step-by-step AI recommendations, and editable outputs for transparency and control.
Prompt DesignGuided UX
4
Scalable Design System
Reusable component library purpose-built for AI and data-heavy UI, supporting rapid feature deployment across all modules.
Design SystemsComponents
5
AI Governance & Guidelines Framework
Internal AI interaction standards: prompt consistency, ethical AI patterns, transparency patterns, and guardrails for decision-critical workflows.
AI GovernanceResponsible AI
Impact

Design that shipped and delivered.

Measured outcomes from enterprise usability testing, production deployment, and client feedback — validated with real finance teams before release.

32%
Improvement in financial processing efficiency
18%
Revenue growth contribution through workflow design
7
Enterprise clients onboarded including 1 unicorn
4
Platform modules shipped: Spending, Procurement, Compliance, HR
↓
Reduction in manual workflow dependency
↑
Executive decision-making speed increased
Key Takeaways

What this project taught me about enterprise AI design.

AI UX must prioritise trust over automation. Users will resist AI that feels uncontrollable — transparency and editable outputs are non-negotiable.
Structured AI workflows outperform open AI interfaces. Pre-defined prompts and guided recommendations consistently outperformed open-ended chat for finance tasks.
Design systems are how AI products scale. Without a robust component library, shipping AI features across four modules would be impossible at enterprise speed.
Enterprise UX is about balancing complexity with clarity. CFOs need the right signal at the right moment, with the ability to drill down when needed.

Interested in working together?

Open to senior AI product design roles, contract work, and consulting.

Get in Touch View Full Portfolio