Flexport has taken a direct approach to closing the gap between fast-moving AI tools and the people who run global supply chains. In January 2025 the company’s vice president of engineering, Alex Nederlof, launched a 90-day internal course designed to teach employees from operations, human resources, legal and other departments how to build their own automated tools. Participants do not write traditional code. They prompt large language models such as OpenAI’s ChatGPT and Anthropic’s Claude, along with AI agents, to generate the software they need. The method has earned the shorthand term vibe coding.
Employees must commit to automating one specific piece of their own workflow. That requirement keeps the training grounded in daily work rather than abstract theory. By the end of the program they learn to think like product engineers. They consider security, testing, human oversight and system reliability. The course has since expanded to two levels and continues to evolve as new automation platforms such as n8n and GitHub features appear.
And the results have been immediate. One senior program manager in operations, Jenna Ward, entered the program with zero coding experience. Months later she had built automations that now drive companywide initiatives. “I went from zero coding experience to building automations that are shaping companywide initiatives — in a matter of months,” she said, according to Business Insider.
Flexport’s broader technology strategy provides the backdrop. The freight-forwarding company has shifted its platform from simple visibility into an automation and execution engine. Its 2026 Winter Technology Release introduced a fleet of AI agents that audit past customs filings for errors, calculate tariff refunds, optimize container utilization to cut freight costs by as much as 10 percent, and translate messages in real time. One internal AI auditor now checks 100 percent of customs entries and has driven the company’s error rate down to 0.2 percent, roughly ten times better than the industry average. These tools rest on a rich data foundation called Flexport Atlas that models live ocean freight movements across vessels, ports and historical performance.
The internal training program directly supports that external product push. Automations built by course graduates have replaced external software-as-a-service subscriptions, lowering costs. Recent student projects include a tool that automatically reviews customs forms and a Q&A chatbot that answers common employee questions for the human-resources team. Both examples demonstrate how non-technical staff can create production-grade solutions when given the right prompting skills and guardrails.
Nederlof, who also leads Flexport’s AI platform team, has fielded inquiries from other companies eager to copy the model. His guidance is straightforward. Identify people with an aptitude for rapid automation, give them strong technology access, and install guardrails to prevent basic mistakes. “These people live in every part of your organization. Drop your preconceived notion of who might be eligible. From ops to recruitment to finance, I’ve seen success stories everywhere,” he told Business Insider.
But the program also addresses a deeper industry problem. Boston Consulting Group managing director and partner Anastasia Kouvela has noted that tools have developed faster than internal capabilities in logistics. A skills gap in data disciplines, limited trust in advanced algorithms, and weak soft skills for explaining AI decisions continue to slow adoption. Flexport’s course attacks all three by giving employees hands-on experience, concrete results and the vocabulary to discuss their creations with colleagues and customers. Business Insider first reported the details of the program and its outcomes.
Flexport has paired the formal 90-day curriculum with lighter-touch learning options. All employees can access a self-guided content library covering AI, logistics and management topics. A dedicated Slack channel encourages staff to share videos, tutorials and successful workflows. The structured course, however, demands dedicated time each week and manager approval. That combination creates both broad exposure and deep expertise.
The company’s CEO, Ryan Petersen, has made AI development a priority. In recent comments he described declaring a “code red” internally to accelerate agent building for tasks once handled through manual emails, spreadsheets and phone calls. Petersen has also spoken about Flexport’s move beyond its startup phase toward larger enterprise customers and eventual public listing. The Loadstar covered that strategic shift in April 2026, highlighting how AI now sits at the center of operations between shippers and carriers.
Concerns about workforce displacement inevitably surface whenever AI gains ground. Nederlof confronts the question head-on. He tells participants that the skills they acquire will make them more valuable wherever they work next. “I carry the same value as four other operators. I am a four-person team, because of the tools that I have,” he says, echoing a message he shares with his own staff. The program therefore serves as both retention tool and portable career asset.
Flexport is not alone in recognizing the need. Industry observers point out that fluency in technology and AI has become a baseline requirement in supply-chain job postings. Companies that invest early in their people gain the ability to squeeze out inefficiencies, react faster to disruptions and deliver superior service. Yet many organizations still stop at basic awareness training. Flexport’s decision to build a rigorous, evolving curriculum that demands real project delivery sets a higher bar.
The course continues to change. When engineering teams realized that email automation once estimated at six months could now be completed in three days, Nederlof adjusted the curriculum. New levels offer fewer restrictions on project scope and tool choice. That responsiveness mirrors the pace of AI progress itself. Participants learn not just current techniques but how to stay current as models and platforms improve.
So far the program has produced tangible wins inside Flexport and interest from outside. Other logistics firms and technology companies have asked for blueprints. Nederlof’s response emphasizes mindset over specific technology. Find the builders already inside the organization. Give them structure. Let them solve their own problems. The supply chain that emerges is faster, cheaper and more accurate. And the employees who build it carry skills that travel with them.
Flexport’s experiment offers a practical template at a moment when many executives worry about both AI adoption and talent retention. The company has shown that non-engineers can produce production systems when given the right training, tools and expectations. In an industry long reliant on manual coordination across borders, that shift matters. It turns static job descriptions into dynamic capabilities. It reduces dependence on outside vendors. And it prepares people for a future in which the ability to direct AI systems may prove as fundamental as the ability to read a spreadsheet once was.
WebProNews is an iEntry Publication