The Road to Clarity — A Builder's Reflection on AI, Logistics, and the Power of Listening
Reflecting on the journey of building Arise & Shine Transporters, I consider the challenges, surprises, and moments of pride that shaped this AI-powered logistics platform. This is a story of listening — to data, to users, and to the tools that helped bring it to life.
"The Road to Clarity — A Builder's Reflection on AI, Logistics, and the Power of Listening Reflecting on the journey of building Arise & Shine Transporters, I consider the challenges, surprises, and moments of pride that shaped this AI-powered logistics platform. This is a story of listening — to data, to users, and to the tools that helped bring it to life. It’s Sunday, and I find myself looking back at the weeks that shaped Arise & Shine Transporters. This logistics platform for sand and aggregate delivery in East Africa wasn’t just a technical challenge — it was a deeply human one. Building it meant learning to listen not only to the needs of the business owner who asked for it, but also to the trucks, the drivers, and the data that moved through the system in real time.
What was hard? Early on, it was realizing that the business owner didn’t just need a tool — they needed a way to see what was happening on the road. Phone calls and paper records were the only way to track where trucks were, how much fuel was being used, and whether a delivery was on time. That lack of visibility was a problem that had never been solved before, and I had to find a way to make it work.
What surprised me? It was the way the AI copilot — that feature that sits with the admin and gives real-time insights — became the most-used part of the platform. I had built it as a convenience, but the admin began using it like a partner, asking it questions, relying on it to flag anomalies, and even to help with daily decisions. It wasn’t just a tool; it felt like a voice that said, 'Wait — this truck is taking a route that doesn’t match the delivery plan.'
What made me proud? When the first truck’s fuel log showed up on the dashboard, and the system automatically flagged a discrepancy between the expected fuel consumption and the actual usage. That moment was a quiet one, but it meant everything. It meant the system was working, and it was working for real people. It was proof that AI, when used with care, could help solve problems that had been ignored for years.
There were days when the code didn’t behave as expected. I remember one instance when the dashboard would freeze if it tried to load too many telemetry rows. It was a small detail, but it was enough to stop the system from working for the user. I spent hours debugging, trying to find the right balance between performance and accuracy. In the end, it was the data that pointed the way — not the AI, but the numbers that told me what was wrong and what needed fixing.
AI wasn’t just a tool in this process — it was a collaborator. I used Claude to think through the logic of the planning assistant, and Codex to help structure the backend code. They didn’t solve everything, but they helped me see the gaps and find solutions that I wouldn’t have found alone. It was a partnership that made the process faster, but it was still my responsibility to build something that worked for real people.
Looking back, I think the most important lesson I learned was the power of listening. Whether it was listening to the data that showed a truck’s route didn’t match the plan, or listening to the users who needed more than just a dashboard — it was the act
