An Aivell answers "where is my shipment" emails from your own tracking data, reads proof-of-delivery photos — even the crumpled ones — matches them to shipments, and chases carriers for the missing paperwork. It logs into carrier portals like a dispatcher would, and everything runs on a box in your office.
Every customer with a truck on the road emails twice a day. Each answer means opening the TMS, finding the load, checking the last carrier update — five minutes of dispatcher time per email, hundreds of times a week.
You can't invoice without the proof of delivery, and the POD is a photo of a crumpled CMR taken in a truck cab at dusk. Collecting, deciphering and filing them is a job in itself; chasing carriers for the missing ones is another.
Five carrier portals, five logins, the same ritual before coffee — download status reports, pull the PODs that finally appeared, re-key what changed. Nobody's best hour, spent copy-pasting.
The carrier confirmed €890, the invoice says €960, and the difference hides in a fuel surcharge line. Catching it means comparing two PDFs against a spreadsheet, load by load.
Freight forwarding has two workdays inside every day: the one spent moving loads, and the one spent documenting that loads moved. Status emails, POD collection, portal rounds, rate-versus-invoice checks — repetitive tasks of exactly the kind that eat 62% of employee time across industries, except in freight the paperwork has a deadline attached: no POD, no invoice, no cash.
And the raw material is hostile. The proof of delivery is a photo of a consignment note that spent four days on a dashboard. The status update is buried in a carrier portal that logs you out every twenty minutes. The discrepancy is one surcharge line across two hundred invoices.
An Aivell is a box in your office that does the second workday for you. It answers the “where’s my container” emails from your own tracking sheet, in the customer’s language. It reads the POD photos, matches them to loads, flags the ones with damage remarks. At night, its Browser Agent makes the portal round you used to make at 8 AM.
You brief it like a dispatcher, in plain language — which carriers get chased and how hard, what a complete POD looks like, when Elena needs to see something before it’s invoiced. Guardrails hold it to the script: it can file and chase, but it can’t rebook a load or invent a status.
And because inference runs onboard, your rate sheets, your margins and your customer lanes are processed behind your firewall — not uploaded to whichever AI vendor your competitor also uses.
every evening, list yesterday's delivered shipments with no pod on file
read pod photos arriving by email or in the scans folder, even the rough ones
match each pod to its shipment number, file it, mark the load ready to invoice
signature missing or damage noted? flag it to Elena before anyone invoices
carriers missing pods older than three days get a chaser — polite the first time
after a second chaser with no answer, put the carrier on friday's escalation list
update the pod tracker so dispatch sees at a glance what's still open
Delivery notes, meter dials and site damage read from photos by a vision model on your own box — and written straight into your database as structured records.
The Browser Agent logs into supplier, bank and utility portals, downloads the month's invoices and files them on your share — passwords never leave the box.
An on-premise AI colleague chases unanswered emails and missing client documents — politely, on a schedule, stopping the moment the reply arrives.
Purchase orders arriving by email — PDF, scan, photo or plain text — read onboard, checked against your prices, entered into your system, confirmed back.
Your info@ mailbox, triaged by a private AI colleague — classified, routed, and the routine ones answered. Behind your firewall, around the clock.
A webhook fires in one system, your Aivell writes the record correctly into the other — on-premise glue between your tools, with guardrails on every insert.
Your Aivell compares spreadsheets with the database, explains every mismatch in one line, and mails the list — financial data checked behind your firewall.
Nowhere. The Aivell runs inference onboard, inside your network — the rate confirmations, the margins, the customer list it reads to do its work are never sent to a cloud AI. In a market where a leaked rate sheet costs you the lane, that's not a detail.
Neither. The Aivell works around it the way a dispatcher does — reading the emails, filing the PODs, keeping the tracker sheets current, pushing updates through email or webhooks. Your TMS stays the system of record; the Aivell does the keying and chasing nobody wants.
It reads scans and photos the way a patient clerk would — shipment number, date, signature present or not, remarks about damage. When a document is genuinely illegible, it says so and asks the carrier for a better copy instead of guessing.
The Browser Agent logs in with single-click passkeys and does the round for you — downloads the status report, pulls the new PODs, files them. Guardrails pin it to the portals you've listed; it can download documents but never change a booking.
An Aivell is an on-premise AI colleague: a small box that plugs into your network, runs its own AI onboard and takes the repetitive work off your desk. Prompts, documents, data — nothing ever leaves your office. Unplug the internet: it keeps working.
You describe each task as a chore, in plain language. Your Aivell turns it into a solid, guarded procedure and runs it in the background — with guardrails on every tool, a complete audit trail, and approvals in your hand when you want them. One-time setup, fixed monthly fee. No tokens, no overages.
onboard inference · air-gapped · fixed monthly fee