53-Year-Old Bellows Manufacturer Nabel Adopts Leach Generative AI Advisory Service and Proposes Raising Monthly Fee After Six Months

Nabel Co., Ltd., a long-established bellows manufacturer founded in 1972 with 199 employees in Iga, Mie Prefecture, is expanding from a hardware-centered business into new software-related development areas such as Robot Insight, a robot condition monitoring service. In October 2025, the company introduced Leach’s generative AI advisory service. Six months later, Nabel proposed increasing the monthly fee before waiting for contract renewal. Nabel began with camera bellows and now provides bellows technology for medical devices, aerospace equipment, measuring instruments, and other industrial systems. In recent years, the company has expanded beyond OEM products into customer-focused businesses such as consumable filters, Robot Insight, Robot Flex, and robot sales as a distributor for a Taiwanese manufacturer, gradually broadening its business from hardware toward software-related areas. Before adopting the advisory service, Nabel faced several structural challenges. The company had no dedicated information systems department or full-time IT staff. Many company-wide operations depended on Microsoft Access, including drawing number issuance, progress management, manufacturing instructions, and drawing templates. Because these systems had grown without requirements definition or flowcharts, bugs, inefficiency, and dependence on specific individuals had become persistent issues. The biggest challenge was that Nabel could not judge whether software estimates from external suppliers were reasonable, approving them without a clear basis for evaluating requirements, effort, or cost. Leach provides three main forms of support: Python study sessions, AWS/IoT architecture support, and real-time technical consultation via chat. The common goal is not to outsource implementation, but to help Nabel build internal judgment capability. The Python sessions help the president and younger members learn requirements definition, flowcharts, and software development principles, laying the groundwork for moving away from person-dependent Access systems. The AWS/IoT support provides third-party reviews of AWS cost, security design, and external development estimates for Robot Insight. Chat consultation fills the gap left by the absence of an internal IT department, offering immediate advice for day-to-day technical decisions. For Robot Insight, Leach continuously joins meetings between Nabel and a major telecommunications company, reviewing requirements, costs, and security from a third-party perspective. Nabel says this is especially valuable because IoT security issues on the end-user side can be serious, and having an external expert assess whether the design risks are appropriate provides significant reassurance. Leach also advises on reducing AWS costs, gateway software costs, and both initial and running expenses. Six months after adoption, Nabel has gained a clear benchmark for reviewing external software estimates. For Robot Insight and related external development proposals, Leach reviews requirements, effort, security, and cost from a third-party standpoint, allowing Nabel to compare multiple candidates rather than ordering development without a judgment standard. Some functions are now being reconsidered for in-house software development by Nabel engineer Mr. AN, with substantial cost reductions expected. In parallel, an overseas development partner found by Mr. AN has proposed an alternative with equivalent functionality and an approximately fourfold cost difference, and Nabel is now evaluating the final vendor structure. President Yoshitomo Nagai said that without the advisor, Nabel might have approved the original estimate without understanding its validity and might not have noticed unrealistic requirements. The project could have taken more time, cost more, or even failed to materialize. Leach’s SaaS product Tsukiawase.com has also been used to automate purchase order-related work, helping an initial automation issue begin operating in practice. Nabel proposed increasing the monthly fee because the purchase order automation had begun to work and the company wanted to increase the frequency of study sessions and accelerate the internal adoption of AI and software capabilities. Nagai also told other manufacturing executives that the working population will continue to decline, and indirect work that does not create productivity should be replaced by AI while companies still have enough people. Waiting five or ten years may be too late.