Building Arise & Shine Transporters with Claude and Codex: The Human Side
From frustration to focus, this journey of building the Arise & Shine Transporters logistics platform highlights how AI collaboration helped solve real-world challenges.
From Frustration to Focus: Building Arise & Shine Transporters
The day I started working on what would become Arise & Shine Transporters was one filled with confusion and frustration. My boss had this idea for a logistics management system, but the reality hit me hard when it came time to build something from scratch — all while managing multiple other projects.
I saw the need for an AI-powered solution that could handle order management, fleet telemetry, driver tracking, cost auditing, financial reporting, and even generate daily thought-leadership articles. This was a real problem in East Africa where sand and aggregates supply businesses were still doing things manually over phone calls or paper records — no visibility into truck locations or delivery costs.
How Claude and Codex Helped
I had heard about AI tools like Claude and Codex, but I wasn't sure how they could be practically applied to my project. The first step was finding an open source model compatible with OpenAI’s API: Ollama. This became the foundation for what we would later use as a base model.
With Claude acting as our primary human-AI collaboration partner and Codex helping us clean up text data, we could reason through complex logistics problems. For instance, determining cost structures based on distance was challenging due to varying local costs in Kenya versus other regions where sand and aggregates are sourced. We needed something that wasn’t just a one-size-fits-all solution.
The Challenge of Distance Pricing
One particular issue I faced involved developing the distance pricing algorithm for Arise & Shine Transporters. This required considering all sorts of factors including different rates for various modes of transportation, varying costs between urban and rural areas, as well as fluctuating demand throughout the day or week based on seasonal variations in construction projects.
I would start working with Claude to brainstorm ideas about how we could structure this pricing model so that it remained flexible yet accurate. We iterated back-and-forth through several drafts until we finally settled on something that worked for our specific needs and provided a solid foundation for the system.
The Moment It Worked in Production
It was incredibly rewarding when everything came together — especially when Arise & Shine Transporters went live and users started experiencing real value from their insights. They were able to see up-to-date information about each truck’s location, monitor delivery costs accurately based on distance traveled, track vehicle performance metrics like fuel consumption and route efficiency.
The system was also capable of generating thought-leadership content every day automatically by analyzing data trends in the industry. This not only helped keep users engaged with useful information but it also served as a valuable tool for internal communication among team members who could use this generated material to inform other decisions or reports.
The Human Element
While Arise & Shine Transporters was technically advanced, I never forgot that at its core, we were solving real-world problems faced by actual people. These individuals included the business owners and managers running sand extraction operations in different parts of Kenya who previously had no way of getting up-to-date information on their assets or costs — it was all done manually.
I often found myself reflecting back to this early stage where I felt overwhelmed, only to realize how much progress we were making. Claude and Codex played a crucial role not just as collaborators in development but also by reminding us that the end goal is always about improving lives and creating tools for real people to use effectively."
