What operationally challenges companies like Cabooter
Cabooter Group operates with rail, road, and warehousing across three terminals in the Netherlands and Germany. That means multiple transport modes, multiple countries, and a constant flow of CMR documents, bills of lading, customs declarations, and container planning. The software supporting this needs to connect with terminal planners, TMS systems, customs platforms, and customer portals. This is precisely where things go wrong for many logistics companies: one package does not communicate with another, and staff manually re-enter data that the system fails to pick up automatically.
When is a standard TMS or WMS sufficient for logistics?
A standard transport management system works well as long as your processes fit within the logic the vendor has built in. For a carrier operating domestically with fixed customers and predictable route structures, a generic TMS is perfectly adequate. The challenge starts when you work intermodally, combine multiple transport modes, or have customers who submit documents in ten different formats. At that point, the standard logic no longer holds. Configuration addresses part of the problem, but every adjustment to an off-the-shelf package costs money and time, and makes upgrades riskier. At some point, you are paying more for the workarounds than for the system itself.
Automating document flows in transport: where does the real problem lie?
Most document problems in transport are not a technical problem but a data entry problem. A driver submits a CMR as a photo, a customer sends a packing slip as a PDF, a terminal exports a CSV with non-standard column names. The system recognises none of this automatically, so someone types it in manually. An AI layer on top of the existing system, in the form of document processing, addresses this directly: it reads the document, identifies the type, extracts the relevant fields, and places them in the correct field of the TMS or customs platform. The person reviews and approves; the system handles the data entry. This is exactly what Bonsai's document processing product Dottle does for logistics companies.
Customs declaration automation for freight forwarders: off-the-shelf or custom?
Established software exists for customs declarations, and for many freight forwarders it works well when the declaration flow is predictable. It becomes a different story when you are dealing with unstructured input from customers, a wide variety of commodity codes, or exceptions that require manual handling. An AI layer can help here by pre-processing documents before they are entered into the declaration system. Custom development only becomes the better choice when the declaration process itself is specific enough that no standard package follows the logic correctly, or when you want to integrate the system with terminal data, clearance, and invoicing in one continuous process. At that point, you are not adding a layer on top of an existing system; you are replacing the whole thing.
When is Bonsai relevant for a logistics company?
Bonsai is not a vendor of standard software, nor an implementation partner for existing packages. Bonsai is a Software Operating Partner: we build systems ourselves, in production, with AI built in. That is relevant in two situations. First situation: you want to extend your current TMS, WMS, or customs system with an AI layer for document processing or order entry, without replacing the core system. In that case, we build an AI Worker that operates on top of your existing system. Second situation: your core system is outdated, no longer aligns with your processes, and configuration costs more than it delivers. In that case, we rebuild the system from the ground up, custom-made, with AI built in. In both cases, the client owns the code and the data. No lock-in, no monthly licence fees for something that only half fits.