Akiba
Context & Premise
AkibaOne is an enterprise fintech SaaS platform built to help financial institutions across Africa assess risk, prevent fraud, and automate operational workflows. In markets where traditional credit data is limited, the platform leverages alternative sources like telco data, mobile money activity, and digital behavioral signals to enable smarter decision-making.
My role was to take this incredibly dense, high-stakes financial logic and wrap it in an interface that prioritized clarity, logical structure, and simplicity. The project demanded a multi-layered approach: establishing a foundational design system, designing a high-friction consumer verification flow, and architecting the platform’s crown jewel—a visual workflow automation engine.
Project Overview
ROLE
UI/UX Designer
TIMELINE
Dec 2024 – Apr 2025
PLATFORM
Enterprise Web SaaS
FOCUS
Systems Design, Workflow Automation, Identity Verification
The Problem Statement
Designing for enterprise fintech means designing for zero-margin-of-error. I focused on solving three distinct UX challenges:
- The Logic Bottleneck (The Builder): Financial institutions needed to automate loan origination and compliance checks, but their operational teams couldn’t write code. They needed a way to visually construct complex automation processes without relying on developers.
- High-Friction Onboarding (KYC): Identity verification is the highest drop-off point in fintech. We needed a consumer-facing passport verification flow that maintained strict compliance without frustrating the user and causing abandonment.
- Inconsistent Foundations: To handle dense data across various modules, the platform desperately needed a unified visual identity and design system built to project trust, intelligence, and reliability.
The Process: Systems Thinking & Lean Execution
To tackle a platform of this scale, I couldn’t just start drawing screens. I had to map the system’s architecture first.
Building the Design System
Before tackling the complex workflow logic, I established the visual identity and design system. For an enterprise fintech platform, the UI must
communicate absolute credibility. I developed a structured component library anchored by a trustworthy primary blue. I meticulously mapped out varied button states, exhaustive text field behaviors, Auth Code error states, and KYB (Know Your Business) timeline indicators. This not only ensured visual
consistency across the massive platform but drastically improved our development efficiency and engineering handoff.
Mapping the Decision Engine
To design the Workflow Builder, I had to deeply understand the platform’s core mechanics. I mapped the entire decision engine down to primary
elements: Triggers (the initiating event), Rules (the conditional logic), API Integrations (external data pulls), and Actions (the automated outcome). By translating these abstract technical concepts into visual, modular “blocks,” I created a mental model that a non-technical user could easily grasp.
The Post-Development UX Audit
(A quick note on how I work): Design doesn’t end at handoff. After the engineering team built the initial version of the Workflow Builder, I conducted a rigorous, post-development UX audit. I stress-tested the live interface to ensure edge cases were handled elegantly and that the shipped product matched the designed experience perfectly. (Because handing off a Figma file and praying is not a strategy).
The Solution: The Decision Engine
Instead of forcing producers to adapt to a clunky AI, I completely redesigned the interface to adapt to the producer.
The Workflow Automation Builder
The core of AkibaOne is the Workflow Builder. I designed a visual, node-based interface on a clean dot-grid canvas that allows users to construct
complex decision trees visually.
- Cognitive Load Reduction: To eliminate guesswork, I attached micro-labels directly above the nodes (Starts when, Check if/else, Do this). Users instantly know a block’s function before reading the title.
- Modular Construction & Branching: Users can seamlessly build out their flow by clicking to add elements directly from the searchable right-hand “Actions” panel. The UI handles complex routing beautifully—for example, a Check KYC node splits cleanly into intuitive If True (Call API) and If False (Deny application) branches.
- Pre-Deployment Testing System: Because a faulty financial workflow can cost millions, I designed a robust testing environment. Before hitting “Publish,” users can click “Run now” to simulate inputs and observe how the workflow behaves, preventing broken logic from ever going live.
To effectively communicate this complex logic to stakeholders, I produced and edited a complete video demo (with voiceover) showing the exact user
journey of creating, configuring, and running a workflow from scratch.
Frictionless Passport Verification
While the workflow builder handled the B2B logic, the system also required seamless consumer-facing inputs. I designed the platform’s passport identity verification flow. Knowing that KYC is inherently high-friction, I focused the UI heavily on real-time error prevention. The camera UI features a dynamic bounding box that turns green with a “Hold still” prompt when the document is aligned. I also included a crucial “Check your image” review
step with a Retake button, ensuring users felt confident before final submission, ultimately reducing API failures and user drop-off.
Conclusion & Results
By establishing a strict design system and translating complex conditional logic into a visual interface, the AkibaOne redesign successfully transformed a highly technical backend into an accessible, enterprise-grade tool.
- Operational Efficiency: Empowered non-technical teams to accelerate credit decisioning and standardize their evaluation logic independently.
- Risk Mitigation: The visual testing environment and transparent execution logs ensured institutions could automate their processes safely and remain fully compliant.
- Engineering Velocity: The comprehensive design system significantly reduced UI bugs and accelerated the development lifecycle for future modules.