ZenRead featured on BRIDGE

On October 21, 2025, BRIDGE published our article titled “Tackling the Reality of ‘80% Unstructured Data’ with AI-BPO: Zenport’s Steps Toward Supply Chain DX.” ZenRead is a service that uses AI to automatically read trade-related invoices and shipping documents, then adds a final human review. This achieves both accuracy and efficiency simultaneously—goals that were previously difficult to attain.

Original article

https://thebridge.jp/2025/10/ai-bpo-zenport-supply-chain-dx-unstructured-data

Translation allowed by the media publisher

Tackling the “80% Unstructured Data” Reality with AI-BPO — Zenport’s Steps for Supply Chain DX

Mr. Fumiyuki Ota, CEO of Zenport, who participated in the press briefing.

This article features entrepreneurs who gathered at a meetup co-hosted by BRIDGE and MUFG Innovation Partners. The next event is scheduled for December. Those wishing to be invited or attend the press briefing can register here.

Zenport, provider of the “ZENPORT” data hub to companies with global supply chains, released the AI-BPO service “ZenRead” in September. This service, which supports the automation of import document processing that has long been done manually, is a stepping stone towards the coming era of AI agents. We spoke in detail with CEO Fumiyuki Ota at our magazine’s press briefing.

“ZENPORT” Adoption Accelerates Among Large Enterprises

Zenport provides a data hub that can centrally manage all data handled in the global supply chain. It connects data across the entire supply chain process—from ordering, shipping, transportation, and inventory to production and sales—providing companies with an environment where they can visualize and operate everything on a single screen.

Adoption by large corporations is currently accelerating. According to Mr. Ota, the service is gaining high praise, with utilization expanding to include overseas bases at major food manufacturers, global consumer goods manufacturers, and major logistics companies.

The background to this is the drastic change in the supply chain environment since the COVID-19 pandemic. Demand forecasting has become difficult, and transportation itself has become a risk due to conflicts and tariff issues in various regions, forcing companies to build new supply chains. However, much of this work is still being done in Excel.

ZENPORT automatically captures internal and external data, including master data, Excel files, and emails, allowing for real-time visibility of outstanding orders, delivery delays, and more. Furthermore, its patented automatic data matching function eliminates manual tasks such as transcribing packing lists and checking schedules on shipping company websites, achieving significant labor savings.

The ability to flexibly adapt to each company’s business flow with a customizable screen function is cited as a reason for its support from large enterprises.

“We are being evaluated for our ability to manage the complex organizational connections and data intricacies unique to global supply chains, and adoption by large companies is steadily advancing,” says Mr. Ota.

An AI-BPO Service That Tackles Long-Standing Manual Work

The newly released “ZenRead” is a service that tackles the “final manual tasks” remaining in this environment.

It is said that 80-90% of data handled by companies is unstructured data, such as emails and various documents, and the same issue exists in global supply chains. Particularly serious is the processing of PDF documents that occurs during importation.

Despite standardized processes worldwide for documents like invoices, packing lists, bills of lading (B/L), and certificates of origin, the work of entering this information into a company’s own core system relies on manual labor.

“Importers open their core system, visually check the data received from exporters, and manually type it into Excel. Then they go to the shipping company’s website to check information and type that into Excel as well. They do this endlessly,” says Mr. Ota.

ZenRead is an AI-BPO service that supports the automation of this import document processing.

Through a series of steps—reading with AI-OCR technology, verification and correction by specialized staff, and automatic integration into core systems—it transforms unstructured data into structured data, making it usable for operations.

What’s important is the recognition that, given the 100% accuracy required for corporate data structuring, there are limits to automation with current AI technology. The company ensures accuracy with a Human-in-the-Loop approach. By collaborating with external BPO partners, they also flexibly respond to fluctuations in data volume.

Aiming for Global Expansion with a Palantir-style Model

This AI-BPO service is positioned as the first step towards building future AI agents. Mr. Ota points out that while AI will become capable of customizing UIs and other elements, the structured data in core systems will remain as an “immovable presence.”

“How do we connect the unstructured data incidentally generated through inter-company communication and other processes with the structured data required for each corporate operation? Especially in BtoB, 100% accuracy is required, making it difficult for AI to handle everything. In terms of being the gateway, hub, or platform for that—SaaS, though it might not be SaaS anymore—that part will not die,” says Mr. Ota.

What is crucial is preparing the foundation for AI agents to operate smoothly. The core of the company’s strategy is to “strengthen the groundwork” by clearly defining data structures and providing appropriate instructions.

One of the features of Zenport’s organizational structure is that it has built an international team from its inception. Seventy percent of its employees are foreign talent, and the official company language is English. They actively hire domestic and international personnel familiar with large corporations and the manufacturing industry.

Currently, 70-80% of users within client companies are based in Japan, but the remaining 20-30% are using the service at overseas bases in North America, Europe, China, and Southeast Asia. The company aims for global expansion using a solution engineer model, similar to Palantir.

“There’s talk of Forward Deployed Engineers (FDE), but we aim to provide a service that properly fits each customer while having a standardized system base. By considering the combination of AI and people on that premise, we believe we can dramatically improve both productivity and quality. We want to be able to do that globally,” says Mr. Ota.

In the supply chain domain, data inefficiency has long been accepted as “unavoidable.”

However, with the advancement of AI technology, a realistic method for finally addressing this deep-rooted problem is coming into view. Zenport’s initiative looks to be a touchstone for the entire industry as it moves to the next stage of digitalization.