What Logistics and Manufacturing Taught Me About Building Software

Many believe transitioning into software and AI means starting from scratch. However, Tony Sterlin Cantave found it more like adopting new tools. With over 15 years of experience in logistics and manufacturing, he developed systems for delivery routes, production management, cleanroom equipment, fleet operations, and shipping. Today, he writes Python and JavaScript, evaluates AI output for training, and builds screen management software for Local Flyer Ads.

Several habits from his previous roles continue to influence his work as a developer. Every system is a flow, and the bottleneck determines its speed. In manufacturing, the line only moves as fast as its slowest station. Similarly, in software, when a script or app feels slow, the instinct is to optimize the code being examined.

The better approach is to measure first and identify the one step that actually slows everything down. Production management taught Cantave to look for constraints before making any changes. Cleanroom work relies on written procedures. Improvisation is not an option when servicing equipment in a controlled environment, as small mistakes are costly and hard to rectify.

This habit translated well into programming, where he writes down what the software should do, its inputs, and how it should respond to errors. This serves as a checklist for testing and makes the code easier for others to understand, including future versions of himself. Handoffs present another area where issues arise. Delivery errors often stem from handoffs: scanned packages not scanned at subsequent stops, labels not matching manifests, changed pickup windows without schedule updates.

Software failures also occur at handoffs, such as functions returning dates as text while others expect date objects or API field name changes causing downstream issues. Cantave addresses these problems by checking boundaries, validating incoming and outgoing data, and logging both. Data entry is often a design issue. ERP and SAP integration taught Cantave that bad data usually originates from poorly designed forms.

Fields accepting any text lead to multiple date formats, and missing dropdown options cause users to select the closest wrong choice. When designing screens and tools, including the scheduling aspect of Local Flyer Ads, Cantave aims to make the correct input the easiest and most intuitive option. Clear choices, sensible defaults, and validation that explains expectations are more valuable than cleanup scripts.

Performance in logistics is measured by on-time delivery rates, accuracy, and turnaround. If you cannot quantify it, you cannot manage it. AI evaluation work through Handshake AI follows the same principle. Cantave scores image, video, audio, and interface output against written rubrics rather than relying on personal preferences, as rubrics ensure consistent and explainable judgments.

He applies this practice to his own projects by deciding what "done" means and checking the project against that list before considering it complete. Reliability is paramount in logistics. A delivery route either succeeds or fails entirely. This principle extends to software development, where Cantave prioritizes error handling, retries, clear logs, and alerts that notify of failures before users notice.

If you are transitioning from operations, logistics, or manufacturing to software, you already possess valuable insights into flows, constraints, procedures, and accountability. To bridge the gap, focus on expressing these concepts in code by automating a repetitive or error-prone task from your past experience using a language like Python.

Cantave pursued CompTIA certifications to structure his understanding of foundational skills, covering areas like hardware, networking, security, cloud, and cybersecurity. He is continuing his computer science degree to further develop these fundamentals. While the tools may change, the underlying thought process remains consistent.