Zencargo got AI working in freight forwarding by fixing its data before it automated the work. AI data readiness came first: clean, structured inputs that the system could act on. Josh Clement-Sutcliffe, AI and Automation Product Manager at Zencargo, explained the approach to aibl in April 2026.
The London-based digital freight forwarder first tried the standard advice of automating the simple cases. In a business with dozens of parties per shipment and no shared standards, that stalled. So in August 2025 it rebuilt the programme around three stages. Here is how it worked.
Because the happy path is rare in global freight. A single shipment can involve dozens of parties across several countries, each with its own systems and standards. Communication runs on email, fax and phone, with documents in every format and language. Zencargo’s early automation could not handle that many variables, so the team went back to the data.
The AI and automation just couldn’t figure out the context and what to do given the amount of variables which had to go into those particular processes.
Josh Clement-Sutcliffe, AI and Automation Product Manager, Zencargo
At Zencargo it meant building its own document parser. Luca Scan, trained and fine-tuned on supply chain operations, turns unstructured emails and attachments into structured data. That feeds Luca, an orchestration layer that reads the intent of each message. Luca checks the booking service, then either moves the booking on or asks for missing information.
It let them see the work being done before asking them to rely on it. Attachments went to the right shipment automatically, and the copying and pasting disappeared. Operators could see every suggestion and correction Luca made. The next challenge is noise: sending operators only what is useful, at the right time.
It definitely becomes easier when it starts, or moves away from being theoretical.
Josh Clement-Sutcliffe, AI and Automation Product Manager, Zencargo
It sets the expectation from the top and hires for curiosity. KPIs track AI adoption across departments, and leaders step in when things stall. Candidates may use AI in technical tests, then must explain what is happening under the hood. One product manager taught themselves to code with AI and is now building an internal training academy.
Josh Clement-Sutcliffe is AI and Automation Product Manager at Zencargo, a London-based digital freight forwarder operating across dozens of countries. He rebuilt Zencargo’s AI programme around data foundations in August 2025. Josh Clement-Sutcliffe spoke to aibl in April 2026.
If your automation keeps stalling on messy inputs, this is the sequence to borrow. The full PDF goes deeper than this page, in Josh’s own words.
For COOs, operations leaders, and AI or automation leads in logistics and other complex, document-heavy businesses. Six pages, about a seven-minute read.
