Skip to content
nofeb

Solutions / Finance and AP

Invoice intake for AP agents

For finance teams and ap automation vendors

Accounts payable runs on attachments from a known set of vendors, and the costly failure is a convincing email from someone who is not one of them.

The problem

An AP agent must read PDFs, match them to purchase orders and chase missing documents, while treating every change of bank details as a possible fraud. Finance teams also need to show where the data lives and who touched it.

How it runs on nofeb

  1. Point vendors at an AP inbox; attachments arrive with their extracted text and stay downloadable byte-exact.
  2. Invoices and receipts are extracted into typed metadata as they arrive: vendor, invoice number, total, currency and due date, with any account number masked, and a receipt's merchant, order number and total. For your own document types, define an extraction schema and run it on your own model key; the output is validated against the schema.
  3. Put your vendors' domains on the AP inbox's sender allowlist, and decide what an unlisted sender means: quarantine for review, or reject at the door.
  4. Each message is labelled trusted, known, unknown or suspicious with the reasons, including its authentication results, a look-alike of a vendor's domain, or a display name that does not match the address. A bank-change request from a suspicious sender never reaches the agent as routine mail.
  5. Keep invoice number, amount, PO match and approval state on the thread as typed metadata, and run the inbox at send_with_approval so every outgoing message about payment is seen by a person.
  6. The audit log records every change made through the API, MCP or the console, and every hold, approval and rejection the server made, with the actor, key, target and outcome. Hand auditors a filtered export as JSON Lines.
  7. Export a quarter's AP mail as mbox or EML with its metadata as JSON; delete or archive mail after your retention period, and put a legal hold on an inbox when a dispute needs everything kept.
  8. Each tenant's mail is encrypted with its own keys; deleting them makes it unreadable.

Where nofeb stops

  • nofeb does not pay invoices or connect to your ledger. Your agent posts what it extracted to the systems that do.

Status of everything this relies on

CapabilityWhat it doesStatus
Attachment text Attachments are listed with their extracted text and can be downloaded byte-exact. Shipped
Built-in and custom extractors Receipts and orders, invoices (account numbers masked), calendar invitations and named entities extracted into typed metadata as mail arrives, plus your own extraction schemas run on your own model key, with the output validated against the schema. Shipped
Byte-exact raw messages Download any stored message exactly as it was received. Shipped
Typed metadata on messages, threads and inboxes Merge-patch keys onto a message, a thread or an inbox, and declare a key's type so every write to it is validated. Shipped
Recipient and sender allowlists and blocklists Lists of addresses, domains or wildcards per organization, pod or inbox: who an agent may write to, and whose mail it reads. A blocked recipient is refused on every autonomy tier, a blocked sender's mail is rejected or quarantined, and you choose whether a send outside the allowlist is held for approval or refused. A check call predicts what a send or an inbound message would meet before it happens. Shipped
Sender trust on every message Each inbound message is labelled trusted, known, unknown or suspicious, with the reasons: its authentication results, your sender lists, look-alike domains, display-name spoofing, and whether the inbox has corresponded with the sender before. Shipped
Inbound mail server that checks mail at the door nofeb's own MX receives mail for the shared domain and your custom domains, with STARTTLS. It checks every recipient before accepting any data, verifies SPF, DKIM, ARC and DMARC, and scans for spam and viruses before it answers. Mail that fails a DMARC reject policy, carries a virus or is clearly spam is refused during delivery instead of bounced later; likely spam and DMARC-quarantined mail is filed into Junk. Every message carries a security verdict with its authentication, spam, virus and TLS results. Shipped
Autonomy tiers enforced by the server Every inbox carries a sending policy, from read-only to autonomous, enforced on the server for every surface that sends, so switching from the API to MCP cannot bypass it. Shipped
Human approval of held sends A held send waits with status awaiting_approval until a human approves or rejects it; both steps emit events. The approval queue lists every held send with its recipients (each marked if new to the conversation), the reasons it was held and who submitted it. Approving needs the separate send:approve scope, which keys from agent self-signup never hold, so an agent cannot approve its own send. Shipped
Append-only audit log Every change made through the API, MCP or the console, and every hold, denial, rejection and approval the server makes on its own, recorded with the actor, key, operation, target, outcome and request id. Query it with filters and cursors, or export it as JSON Lines. Shipped
Full export A whole inbox, a folder, a thread or a search as mbox or a zip of EML files, with every message's folders, keywords, metadata and decisions as JSON, built in the background and downloaded through signed links. Shipped
Retention policies and legal holds Delete messages older than a number of days, or move them to Trash or Archive, per organization, pod, inbox or folder, keeping flagged mail by default, with a dry run that shows what a policy would do. A legal hold blocks every deletion in its scope, including expiry and inbox deletion. Shipped
Per-tenant encryption with crypto-shredding Stored mail and metadata are sealed with keys derived per tenant; deleting a tenant's key material makes its data unreadable. Shipped