5 months
Live before a government suspension
Work / Korensi
Korensi
Sierra Leone's government launched Korensi to move public transport from cash to contactless cards and apps. I designed the admin side: the platforms that oversee the whole system and run each transport company.
- Role
- Product designer, admin platforms
- Type
- Web app, 2 platforms
- Team
- 5 designers, 2 PMs
- Status
- Launched Feb 2024, later suspended
The problem
Minibuses, taxis, okadas and kekes all ran on cash. Korensi launched on 2 February 2024 and recorded over a million ticket transactions in its first month.
The customer, driver and agent apps handled transactions, but nothing managed the system behind them. The government needed to oversee several bus companies, track tickets across hundreds of agents, prevent fraud and make decisions from data.
The question: how do we turn complex, multi-layered admin data into interfaces that give the government oversight and give operators independence, with clear permission boundaries?
Research
This was new ground with no system to audit. I studied Nigeria's BRT operations, fleet tools like Uber for Business, city transit dashboards and enterprise admin tools.
Most transport systems either had one admin tool with messy permissions, or separate tools with siloed data. The PMs also flagged the biggest risk: with paper tickets, nobody could tell if an agent sold tickets and kept the cash. That became a primary design driver.
Key decisions
Four decisions that shaped both platforms.
Two platforms, not one
- Decision
- A Super Admin platform for the government and a Group Admin platform for each bus company.
- Why
- Operators could run day to day without waiting for approvals. One company could never see another's finances. Each interface stayed focused.
- Trade-off
- Two codebases to maintain, accepted for clarity and security.
A dashboard that informs
- Problem
- Vehicles, routes, terminals, transactions, wallets and inventory were separate modules. Admins would click around to piece things together.
- What I did
- A home dashboard with system health, transaction trends, route and terminal activity, and links into each module.
- Principle
- One screen, one job. The dashboard informs; actions happen in modules.
Tracking tickets by booklet
- Problem
- There was no reliable way to catch agents selling tickets without remitting revenue.
- What I did
- An inventory module that tracks tickets by booklet, shows in stock versus sold, and looks up any ticket by ID.
- Result
- If an agent got 500 tickets but paid in for 300, the gap showed up straight away. Kept to Super Admin to avoid conflicts of interest.
Modules first, dashboard last
- Approach
- I designed each module first (vehicles, routes, terminals, transactions, inventory), then built the dashboard from what they really held.
- Why
- The dashboard showed real data instead of guesses, and every card linked to the module it summarised.
Patterns across both platforms
Progressive disclosure: summaries on dashboards, detail in modules. A persistent sidebar so nobody gets lost. Lower density on dashboards for scanning, higher density in modules for real work.
Status colours applied the same everywhere (green healthy, yellow attention, red issue, grey pending) and date filters on every analytics view, because admins think in "this week" and "last month".
Results
1M+
Tickets managed through the system
0
Reported issues with the admin platforms
Design ran April to September 2023. The system launched in February 2024 and was suspended by the government in mid-2024 for reasons not shared with the team.
What I learned
Admin tools need clarity, not delight
Success meant quick decisions, visible fraud and tools built for daily power users.
Fraud prevention is a design problem
It meant understanding how fraud happens and making gaps easy to audit, without false alarms.
Ask about decisions, not tasks
With no access to end users, I learned to ask PMs "what are they deciding?" and "what if they miss it?"
Next time
Test earlier with real admins, design for mobile, and add predictive alerts instead of only reporting.
Next project
APil Me
Savings, insurance and pensions for Nigeria's informal sector.
APil Me / Cover
Have something to build?
I take on product design and build projects for founders and teams anywhere. Tell me what you're making.
adesodun67@gmail.com