When “Self-Serve” Still Requires Humans (The Context)
Honeycoin is a B2B payments platform used by PSPs, banks, financial institutions, and IMTOs to move money across multiple markets.
Although the B2B platform enabled self-service flows, many clients still relied on Honeycoin’s internal team to execute trades, manage FX, and reconcile settlements.
This typically involved the CEO/CFO setting up multiple WhatsApp groups for different businesses/operators and manually updating a spreadsheet for reconciliation and audit with other team members.
Behind many client transactions was an ops workflow that needed speed and accuracy.
This meant, behind every client transaction was a human-dependent workflow and as transaction volume grew, friction compounded quietly leading to errors through delays, context loss, and escalating operational risk.
To address these issue, we developed Nexus, an internal tool designed to automate the reconciliation process, improve data accuracy, and ensure enhanced operational efficiency for a process that had been manually operated for five (5) years.
The Ops Workflow Was Built on WhatsApp and Spreadsheets (The Problem)
The issue wasn’t that the team lacked skill. The issue was that the tools (WhatsApp + Spreadsheets) the internal ops team was using for client interactions and trade management weren’t designed to support operational scale.
Trade data lived in one place. Conversations lived somewhere else. Reconciliation logic lived in another spreadsheet entirely. There was no single source of truth.
This worked when volume was low, but at scale, it created real operational risk:
#1 - Scattered data and inconsistent records
Operational risk due to the data and records been scattered across chats.
#2 - No single source of truth
So much time lost chasing context for trades, FX, or client status instead of executing transactions
#3 - Reconciliation Issues
Reconciliation errors and duplicated effort were hard to trace, often inaccurate, and lacked transparency.
#4 - Poor Team Collaboration
Low visibility across team members as handovers came with loss of critical information.
When something went wrong, the team spent more time finding information than fixing the problem. The system didn’t fail dramatically, it failed quietly, through friction, delays, and compounding risk.
.jpg)
The system of record was fragmented, chats for context and spreadsheets for tracking.
How do we solve it → We Needed a System of Record,
Not a Better Spreadsheet
This wasn’t about building an “admin panel”, or just a productivity problem or just plugging automations or a random fix to it. It was a scaling problem. Honeycoin had outgrown manual operations. The company needed an internal system that could:
Every step had legal implications. Regulatory bodies required specific data formats and documents, compliance workflows needed audit trails, and backend systems were tightly coupled to government processes. This meant improvements had to be incremental, production-safe, and legally accurate — even when that conflicted with ideal UX patterns.
Track every trade and payment as a structured record
Give the team shared visibility in real time
Make reconciliation auditable and repeatable
Reduce dependence on one person holding context
With these issues in mind, the question became:...
So...how might we...centralize and automate Honeycoin’s internal B2B operations so the team can reconcile faster, reduce errors, and serve clients with more confidence at scale?
Reusing the B2B Platform’s Design System
Nexus shipped after the B2B dashboard redesign. I intentionally reused the same design language to keep Honeycoin’s ecosystem cohesive:
Table structures and filters
Form patterns and validation
Status badges and color semantics
Drawers and detail views
Layout rhythm and spacing
By extending the B2B design system into Nexus reduced onboarding friction for operators and shortened engineering build cycles. The internal platform felt like a natural extension of the product ecosystem, not an isolated admin interface.
Testing With Real Ops Scenarios
Because Nexus was replacing a five-year manual system, traditional usability testing wasn’t enough.This wasn’t a greenfield prototype. It was infrastructure migration.
Instead of static test cases, we validated the system against real operational scenarios, the exact workflows the team executed daily under time pressure. We walked through:
A new trade request from initiation to settlement
A trade exception and manual correction flow
FX rate updates and liquidity checks during peak periods
Client onboarding and transaction lookup
Audit-style review of historical records and status transitions
Each scenario was run with operators, not observers. We documented breakdown points, ambiguous states, missing context, and friction in handoffs.
Validated with real workflows, including
exceptions and audit needs.
The goal was simple: If Nexus couldn’t survive real operational pressure, it wouldn’t ship. This process surfaced subtle issues early particularly around state logic and exception handling, and allowed us to refine the system before release.
As a result, Nexus didn’t feel like a new tool. It felt like the natural evolution of how Honeycoin already worked.
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.avif)
.avif)
.png){kind=avatar}
