Company
Shufti โ global KYC/AML identity verification platform
SHUFTI PRODUCT TRANSFORMATION ยท ENTERPRISE DECISION JOURNEY
How I restructured Shuftiโs enterprise self-service journey across onboarding, verification, pricing, review and continuation

Understand. Choose. Review. Confirm. Continue.
Shufti is a global KYC/AML identity verification platform serving enterprise customers across 200+ countries. As Senior UI/UX Designer, I owned the redesign of the self-service decision journey โ the flow where an enterprise customer configures verification, understands pricing, reviews their choices and commits.
The core design idea: make each decision understandable before asking for commitment. I prioritised visible consequences, review and recovery over the shortest possible flow.
Shufti โ global KYC/AML identity verification platform
2026 ยท ongoing product transformation programme
Senior UI/UX Designer
Enterprise self-service onboarding, pricing, review and continuation
Campaign, PPC and acquisition continuity around the decision journey
Embedded with product, engineering, marketing, content, compliance, SEO, sales and leadership โ I owned the design workstream end to end
2026 ยท ongoing product transformation programme
Figma (design, prototyping, QA) ยท Figma AI ยท Claude + reusable Claude Skills
Designing Local Depth at Global Scale โ the global country-page experience system
This role required balancing product UX, growth design and enterprise system thinking across multiple teams and surfaces:
Identity verification is not a routine sign-up category. A customer evaluating Shufti has to work out which checks their organisation needs, how those checks affect their own users, how the service will be priced, and what happens after the configuration is submitted. These decisions influence conversion, fraud exposure, compliance and implementation effort.
The self-service platform connected several of those decisions โ business onboarding, verification pathways and pricing. But the journey did not begin or end inside the platform. A customer might arrive from a campaign page, move through a country or product page, compare pricing, and only then enter self-service. Each surface had grown for a different purpose, yet the customer experienced them as one relationship with the company.
Shufti needed customers to make high-stakes decisions without losing confidence as they moved between discovery, evaluation, pricing and onboarding.

The initial assumption was that enterprise users were struggling because the experience contained too much complexity. The obvious fix โ fewer screens, more hidden information, a compressed journey โ would have made the interface look simpler without making the decisions easier.
Mapping the connected journey across acquisition, pricing and onboarding showed that three questions were repeatedly left unanswered:
What does my organisation actually need?
What will this decision affect?
What happens after I continue?
That revealed the useful distinction between system complexity and decision uncertainty. The platformโs genuine complexity could not be removed โ different organisations need different verification, compliance, deployment and pricing configurations. But uncertainty could be reduced through sequence, hierarchy, context, review and feedback.
Uncertainty was not separate from friction. It was a form of friction. The goal changed from the shortest journey to the clearest progression.
Before committing, I compared three representative structures built from the same content and constraints:
Business info, verification and pricing merged into fewer screens. Fewer steps, but consequential decisions competed for attention at once. Rejected.
Visually sparse screens with little supporting context. This risked moving customers forward without enough information to judge the consequences. Rejected.
Context โ requirements โ pricing โ review โ continuation. More visible stages, but each stage prepared the next decision. Selected.
I judged the three against five criteria: decision visibility, consequence clarity, recoverability, content scalability and implementation realism. I selected the progressive structure for its balance of clear decisions, recoverability and implementation feasibility, accepting a higher screen count.

The final journey moves from broad context to specific commitment in five deliberate stages:
Identify the organisation and the purpose of the configuration first, with help text sitting beside the fields it explains โ not in a distant help area.
Introduce verification choices only after enough context exists to make them meaningful. One dominant decision per stage; supporting information clearly secondary.
Treat pricing as part of the decision, not an isolated commercial page. The right commercial context appears next to the choice it informs.
One reviewable summary organised around the customerโs decisions, not a field-by-field replay of the form. Editing an earlier choice never destroys progress.
The final state marks the stage complete and makes the handover to sales, implementation or account management explicit.
Understand. Choose. Review. Confirm. Continue.
The trade-off was accepted deliberately: more visible stages than a compressed form. A journey with fewer screens can still feel difficult when customers pause, reread, backtrack, or continue without understanding consequences. Every stage had to earn its place โ it stayed only if it clarified a decision, reduced uncertainty or prepared the next action.

The final structure was not a first draft. The journey went through repeated design-review cycles with product, sales and compliance, and each round changed real decisions: the review state was reorganised around the customerโs choices after early versions replayed the form field by field; pricing context was moved closer to the configuration after sales feedback showed customers committing without understanding commercial effect; and stage boundaries were redrawn when a stage failed the "does it earn its place" test. Formal usability testing was unavailable, so structured critique and cross-functional feedback guided iteration.
For the portfolio, I reconstructed the decision flow using authentic product screens. The sequence explains the movement from business context to continuation, including review, editing and recovery states. This reconstruction is separate from the shipped product.
The interaction specifications cover visible keyboard focus, validation with labels and supporting text, errors next to the affected input, and loading feedback. Colour is supported by text, long headings can wrap, and mobile layouts reorganise the decision sequence. These are design requirements, not a claim of independently verified accessibility conformance.

The customer journey started long before self-service. Campaign pages, PPC landing pages, country pages, product pages and pricing all shaped how customers understood Shufti before onboarding. These surfaces did not need to look identical โ they needed to behave like parts of the same organisation.
I extended the same interaction principles across the acquisition journey: a shared action hierarchy, consistent form behaviour, trust evidence placed next to the claims it supports, and product information organised around customer questions instead of an undifferentiated catalogue. Marketing pages kept their narrative freedom; product surfaces kept their functional density. The aim was continuity, not uniformity.

Shipped a restructured enterprise decision journey โ business context, verification, pricing, review, recovery and continuation now behave as one connected sequence instead of a fragmented form.
Connected configuration decisions to contextual pricing, editable review and clear next steps.
Designed structured decision stages, explicit review and safe recovery to address potential onboarding drop-off. The effect on completion rates remains unmeasured.
Unified acquisition, pricing and onboarding under one interaction language, with shared patterns for actions, forms and transitions between product and marketing surfaces.
Delivered a complete, documented interaction state model โ focus, validation, error, loading, review, edit and continuation โ that engineering could implement and maintain consistently.

The work crossed product, engineering, marketing, content, compliance, SEO, sales and leadership โ each with different priorities. Journey maps, alternative flows, prototypes, component specs, responsive states and design reviews became shared decision-making artefacts: a request to shorten the journey could be tested against decision clarity, a visual variation against the shared interaction language. I also kept design completion separate from live implementation โ a finished Figma file was never presented as shipped until the live experience could be verified.
AI accelerated the operation without defining the strategy. I used Figma AI and Claude for early exploration, repeated page structures, diagrams and recurring production tasks, and converted repeated instructions into reusable Claude Skills carrying visual rules and review criteria. AI generated options and first passes; I owned product logic, hierarchy, accuracy, accessibility and approval.
Connect acquisition, pricing, onboarding and activation analytics into one funnel, then measure more than completion: can customers explain what they selected, identify what it affects, return to an earlier choice without losing context, and accurately describe the next step. That is what moves this work from demonstrated structure to measured performance.
This project changed how I define scale. Scale is not the number of screens a designer can produce โ it is the number of future decisions a system can support without losing its purpose. At Shufti, my role moved beyond individual interfaces: I was designing the rules through which customers, teams, markets and production systems could move forward with confidence.
What I would do differently: push for lightweight funnel instrumentation and moderated usability sessions before the redesign, not after โ the structural argument is strong, but a baseline would have let me pair every design decision with measured behaviour from day one.
Do not remove complexity that users genuinely need. Design the confidence required to move through it.