Capstone Trucking
A 19-truck long-haul fleet running on paper forms and Excel needed a real system — one that works in the bush with no signal, surfaces maintenance risks before they become breakdowns, and gives every role exactly what they need.
19
Trucks managed
Full fleet visibility in real time
207
Git commits
6-week intensive build
4
User roles
Driver · Supervisor · Workshop · Admin
AI
Maintenance alerts
Anomaly detection on mileage & behaviour
The Challenge
Capstone ran their entire fleet on paper — driver inspection forms, job cards, maintenance records. Supervisors chased drivers on WhatsApp to find out where trucks were. Workshop decisions relied on whoever remembered what work had been done last.
Long-haul routes through Southern Africa mean trucks spend hours in areas with no mobile data. Any digital solution had to work completely offline — or it wouldn't be adopted.
What We Built
A mobile-first fleet management platform built specifically for Southern African conditions. The React Native app runs fully offline using SQLite, syncing automatically when connectivity returns. No data is ever lost.
On top of the operational layer, a Claude-powered AI agent analyses every report — spotting patterns in mileage, fuel use, and driver behaviour that indicate maintenance needs before a breakdown forces the issue.
Built for Southern African realities
We designed the offline-first architecture from day one — not as an afterthought. Drivers in the Limpopo valley with zero signal can complete their full daily inspection, and everything syncs the moment they hit a town. This was a core requirement that shaped every technical decision.
Platform Features
Mobile-First Driver App
Drivers complete daily vehicle inspection forms on their phones — replacing paper checklists entirely. Forms are designed for one-handed use on the road, with large touch targets and minimal text input.
Offline Sync
The app works fully offline using local SQLite storage. When connectivity returns — at a depot, truck stop, or town — everything syncs automatically. No data is lost when drivers are in remote areas with no signal.
AI Maintenance Manager
A Claude-powered agent analyses incoming driver reports and maintenance logs, flags anomalies in mileage patterns and driver behaviour, and raises prioritised maintenance alerts before breakdowns happen.
Workshop Job Cards
Workshop managers receive digital job cards directly from the AI maintenance agent. Parts usage, labour time, and job status are tracked in the system — replacing the paper job book.
Real-Time Fleet Dashboard
Supervisors and admin staff see the full fleet status on a live dashboard — which trucks are on route, which are in workshop, outstanding driver reports, and AI-flagged risks.
Multi-Role Authentication
Four distinct user roles — each with their own view and permissions. Drivers only see their truck. Workshop staff see job cards. Supervisors see the fleet. Admins see everything.
Tech Stack
Mobile App
React Native · Expo
Frontend
React · TypeScript
Backend
Flask · Python
Database
Neon Postgres · SQLite (offline)
AI
Claude AI (Anthropic)
Auth
Firebase Authentication
Deploy
Railway (backend)
Running a fleet or logistics operation?
Let's digitise your operation.
Whether it's 5 trucks or 500, we'll design a platform around how your team actually works — built for the realities of Southern African transport.
Start a ConversationLet's talk about your
next platform.
No forms. No lengthy RFPs. Start with a conversation — tell us about your business and we'll show you exactly where AI integration creates the most value.
Based in UK · Serving Zimbabwe & Southern Africa · We typically respond within 2 hours