Employees email a photo of the receipt; an Aivell reads it on the box — amount, date, merchant, VAT — categorizes the spend against your policy, flags anything over the limits, and books one clean row per receipt into the expense sheet. Receipt photos never leave your network.
starts when: An employee emails a photo of a receipt to expenses@
Expenses are the paperwork everyone postpones. Receipts live in wallets, jacket pockets and camera rolls until the month closes — then someone in accounting spends an afternoon squinting at faded thermal paper, retyping amounts and arguing about whether a client dinner was “meals” or “entertainment”. Admin like this quietly costs companies around $17,000 per employee per year.
Cloud expense apps solved the typing and created a different problem: every receipt photo gets uploaded to somebody else’s servers. Receipts say a lot — who travelled where, who ate with whom, what was bought at the pharmacy. Plenty of firms would rather that stayed in the building.
The rules that make this work — limits, categories, who approves what — are yours, written in plain language. The guardrails do the rest: the Aivell can add expense rows but cannot touch past months, and it can only ever write to the person who sent the receipt.
No flowcharts, no code — a chore is just a message to your Aivell, in your own words. It shapes it into a solid, guarded procedure and follows it to the letter. This one:
watch expenses@ for receipt photos and pdfs
read each receipt — amount, date, merchant, vat
categorize it — travel, meals, supplies, other
meals over €50 a head get flagged for Paolo
unreadable total or missing date? ask the sender for a better photo
same merchant, same amount, same day as another receipt — flag it, don't book it
add one row per receipt to the expense sheet, tagged by employee
on the 1st, send each employee their monthly recap
Connected with single-click passkeys — no copied tokens, no OAuth hell.
Every tool can be limited to exactly what this chore needs. For this one, you might set:
Mostly, yes. The vision model on the box reads thermal paper going grey, folded corners, handwritten tips and phone photos taken at an angle. When something genuinely cannot be read, it does not invent a number — it emails the sender and asks for a sharper shot.
Only your box and your people. The image is read by a vision model running inside your office and then filed with the expense row. Receipts can be surprisingly personal — where someone travelled, what they bought at the pharmacy — and none of it goes to any cloud service.
Yes — the policy is the chore. Per-head meal limits, categories, which spend needs a named approver, what happens with duplicates. You write it the way you would brief a new bookkeeper, and change it by editing a sentence.
A rule in the chore. Same merchant, same amount, same day gets flagged instead of booked, and a human decides whether it is a duplicate or two genuinely identical coffees.
An on-premise AI colleague reads every supplier invoice — PDF, scan or photo — and books it into your sheet or database. Invoices never leave your office.
Expense, discount and purchase requests arrive pre-checked: case summarized, policy applied, one tap to decide. All prepared on your own hardware.
Hours chased, checked and compiled into one master sheet before the payroll cutoff — by an AI colleague that works entirely inside your office network.
An on-premise AI colleague for accounting firms: reads invoices, chases client documents, minds the tax calendar. Client files never leave your office.
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