- Create orderdoneWien → Hamburg · 8
I prepared the transport details. Review the route, cargo and missing information before approval.
AI agents & Copilot
Eight agents that work with you.
And a Copilot that knows everything.
Each agent specialises in one task and works within its autonomy level. The Copilot answers questions about all your data and carries out actions – strictly within your permissions.
LogCon · Transport file
AI agents · activity
- Mail agentL2
- Document agentL3
- Dispatch agentL1
- ETA agentL3
- Customs agentL2
- Billing agentL2
Which tours to Italy tomorrow are not dispatched yet?
Copilot3 orders to Italy are open. GU-481 LC fits 2 of them – shall I dispatch them as a proposal?
Transport workflow illustration
The agents
Mail agent
Classifies e-mails, extracts orders, requests missing details, answers status questions.
Document agent
Recognises CMRs, PODs, invoices and delivery notes and attaches them to the right file.
Pricing agent
Calculates prices from customer rates and creates spot quotes.
Dispatch agent
Ranks vehicles and carriers by route, driving bans, driving hours and cost.
ETA agent
Calculates arrival times from position, route and EU driving-time rules and reports delays.
Customs agent
Suggests HS codes, checks documents and sanctions lists.
Billing agent
Creates invoices after the POD and checks incoming carrier invoices.
Control tower agent
Finds exceptions – delay, missing POD, customs hold, margin outlier – and proposes actions.
Copilot
Chat assistant in every view: understands context, answers questions, carries out actions as proposals.
How an agent works
- 01
Trigger
An e-mail, a document, a status change or a schedule starts the agent.
- 02
Proposal
The agent produces a structured proposal with confidence, reasoning and risks.
- 03
Autonomy level
The policy decides: log only, suggest, submit for approval or execute.
- 04
Log
Every execution lands in the audit log with before/after state – and can be undone.
Reliable, not magic
- Structured output instead of free text
- Confidence per field and required-field checks
- Transparent AI cost per run
- Multilingual communication with AI disclosure
- Copilot limited to the user's permissions
- Demo and test mode without an AI key
Frequently asked questions
Which AI models does LogCon use?
The agents run on Claude models by Anthropic. Fast models handle simple classification, more capable ones handle extraction and dispatch. The model provider is disclosed as a sub-processor.
Is our data used for training?
No. Your data is processed solely to carry out your orders.
Can the Copilot change data?
Yes, but only within the permissions of the signed-in user – and changes run as proposals with an audit entry.
Copilot
Your transport-file Copilot
Find orders, understand operations and prepare changes using the context of your current page and tools available to your role.
Relevant workflows: Review transport details, find orders and prepare dispatch proposals.
Interactive simulation · no customer data, no sending
Demo · Dispatch
Copilot
Prepare an order Vienna → Hamburg for 8 pallets.
Prepare a change
PendingMail agentCreate orderL2 – Approve
86%
Reasoning: Route and cargo come from this fictional request. Loading time, customer and price still require review.
RouteAT Wien → DE Hamburg2 stops
Shipments1 shipment
Transport workflow illustration
Available tools depend on role, plan and configuration. Copilot mutations prepare proposals. Background agents separately follow configured autonomy levels.