01 · In-App AI
It assists your teams in each module
Extraction from quotes, invoices, transport documents, rate sheets and attachments. AI removes re-entry and prepares structured data.
The AI-ready TMS for freight operations
AI built from document to action
OneChain separates three levels: extraction embedded in modules, Zach for selected email flows, then conversational and headless capabilities whose access still needs qualification.
Discuss your AI roadmapOneChain AI stack
The 3 AI layers
In-App AI Read and extract
In-Product AI To be validated
External AI Zach + access to qualify
AI architecture
Three levels from document to action
The point is not only to read an invoice. The point is to make transport data usable, queryable and actionable.
02 · In-Product vision
An assistant for operational questions
The vision is to let teams query OneChain about risky shipments, costs, delays, missing documents or invoice discrepancies. Its status remains to be confirmed.
03 · External AI
It automates with your rules and guardrails
Zach operates on configured email flows. Opening OneChain to your own agents through API, CLI or MCP still needs qualification with the product team.
Two visions 
Zach reads forwarder emails, classifies documents and updates shipment files. Headless aims to expose OneChain data and selected actions to your internal agents. Planned guardrails: organization-level scopes, read-only by default and explicit approval for sensitive actions.
Zach to operate, Headless as an access program
Zach covers configured incoming flows. The Headless platform describes a second orchestration model to qualify with data or AI teams.
Zach reads forwarder emails, classifies documents and updates shipment files. Headless aims to expose OneChain data and selected actions to your internal agents. Planned guardrails: organization-level scopes, read-only by default and explicit approval for sensitive actions. Option A
Agents operated by OneChain
A shared email address receives forwarder messages. Zach matches attachments and surfaces exceptions.
Option B
Agents operated by the customer
In the target model, your internal agent would query OneChain, reconcile ERP data and prepare operational decisions.
Shared foundation
A TMS as reliable source
The same files, statuses, costs, documents and business rules already feed teams and dashboards. They would also provide context to authorized agents.
Use cases
AI serving business modules
Every AI capability is tied to a concrete freight flow, not to a generic assistant.
Transport
Attach documents to the right shipment
Booking confirmations, HBL, HAWB, POD, CMR and forwarder messages are read, classified and matched to the right shipment.
See transport managementProcurement
Turn offers into comparable lines
PDF, spreadsheet or email quotes become a comparable base for RFQs, spot and negotiated rate sheets.
See rate managementInvoice control
Detect discrepancies and prepare action
AI matches invoice, quote and rate sheet, then prepares an alert your teams can act on.
See invoice controlReporting
Query costs, lead times, CO₂ and quality
Consolidated freight data becomes usable for performance questions and management summaries.
See reportingZach
Connect the TMS to the inbox
Zach reads incoming flows, reduces email loops and only flags exceptions that require a decision.
See ZachHeadless
Connect your own agents
The API, CLI and MCP vision must be qualified with IT, data and AI teams before any technical opening.
See Headless Operational proof
AI anchored in real freight flows
Public and anonymized cases show the value of the foundation: fewer emails, less re-entry, and data that can be used to control cost and execution.
8,000 emails avoided per year at Christofle
Invoices discrepancies matched to quotes and rates
Shipments documents attached to freight context
Evaluate OneChain
Request a free demo Ready to save 50% of your transport management time?
See in 30 minutes how OneChain adapts to your flows, providers and quality standards.