Questions buyers ask about Pypelyne.

What is Pypelyne?

Pypelyne is the autonomous invoice pipeline for Oracle Fusion. Email comes in, validated FBDI goes out — no AP user interface, no review queue, no humans in the loop on the happy path. AI handles the chase when something's missing.

What does "autonomous" actually mean?

There's no operator screen. Suppliers email invoices, the pipeline extracts and validates them against your Fusion data, and a clean FBDI file lands in OIC. If something doesn't validate — wrong PO number, ABN mismatch, line totals don't tie — the AI emails the supplier and works it out, then completes the same flow.

You're not in the loop unless the pipeline genuinely can't resolve it.

How is Pypelyne different from Coupa, Bill.com, or Tipalti?

Different category. Those products give your team a better workspace — better OCR, better routing, better dashboards. They still assume humans inspect and approve invoices in their UI. Pypelyne assumes the opposite: the only good invoice screen is the one nobody has to look at. We sell a pipe, not a workspace.

Who is Pypelyne built for?

Enterprise IT teams running Oracle Fusion. Specifically, organisations where invoice volume creates real operational cost on the inbound side and AP teams would prefer to spend their judgement on exceptions, not on data entry.

We don't run Oracle Fusion. Is Pypelyne for us?

Not yet. Pypelyne is purpose-built for Fusion's FBDI ingestion model. We're not generalising to other ERPs — that scope discipline is intentional.

Who in our organisation should be evaluating Pypelyne?

Pypelyne is a platform service, not a finance product — it integrates directly into your Oracle integration layer (OIC, FBDI), and there's no UI for AP staff to operate.

The natural sponsor is enterprise IT: CIO office, platform engineering, integration team, or whoever owns Fusion configuration. AP leadership is a stakeholder in the outcome — they get capacity back — but the technical evaluation sits on the IT side.

Which Oracle Fusion versions are supported?

Pypelyne is designed against Oracle's current published FBDI and OIC integration patterns for Cloud Fusion. Specifics are best confirmed against your environment on a call.

Do you use FBDI or REST?

FBDI for invoice load — it's the documented bulk-load path. REST and OIC for lookups during validation: suppliers, POs, receipts, tax codes.

Does Pypelyne run on OCI?

Yes — Pypelyne is OCI-native and in-region. That matches Fusion customers' data residency expectations and avoids the cross-cloud data flow most third-party AP vendors create.

What if the AI gets it wrong?

Two gates catch errors. Pypelyne's own validation refuses to emit an FBDI file that doesn't tie. Fusion's FBDI processor refuses anything that fails its rules. Anything that fails either gate goes back to the supplier through the chase loop, never to your AP team. Every decision is logged with the source data.

What does "the chase" mean?

When the pipeline can't validate an invoice — wrong PO, ABN doesn't match the supplier record, line totals don't tie to the header — the AI composes a polite, contextual email to the supplier with the specific missing or wrong data, sends it from your AP inbox, and waits for the response.

When the supplier replies, the pipeline picks it up and runs again. No human intervention unless the loop genuinely stalls.

How long does the chase run before escalating?

Configurable. Default is three attempts over five business days. After that, the invoice still posts to Fusion with a hold flag and the full chase history attached — Oracle's existing exception workflow takes it from there. Pypelyne doesn't hold stuck invoices.

What's the audit trail?

Every invoice has a complete trace: original email and attachments, every extraction decision, every validation result, every chase email sent and received, every FBDI file emitted. Stored in your OCI tenancy or ours — your call. Retention is configurable.

How long does implementation take?

Faster than enterprise AP rollouts. Meaningful estimates require knowing your supplier count, OIC posture, and existing email handling — that's a discovery-call conversation.

Who builds the OIC integration?

We do. We deliver the OIC artifacts — connections, integrations, lookups — into your tenancy. You don't need an Oracle implementation partner to stand up Pypelyne. That's a structural choice — it removes a dependency that adds months to most enterprise rollouts.

Who owns the OIC artifacts?

You do. They're deployed into your tenancy and you have full visibility. If you stopped using Pypelyne, you'd retain the integration code.

Where does the inbound email land?

Your existing AP inbox, or a new one — your call. Pypelyne reads the inbox via Microsoft Graph (Microsoft 365), Gmail API (Google Workspace), or IMAP for everything else. We never ask you to change supplier instructions or migrate inboxes.

What change management is involved with our suppliers?

None on day one. Suppliers keep emailing the same address, and the chase loop replies from that same address — Pypelyne is invisible to them. Routine queries get answered on the pipeline's timescale rather than queueing for a person, freeing your team to focus on the exceptions that genuinely need human judgement.

How much does Pypelyne cost?

Pricing is per validated invoice posted to Oracle Fusion, with volume tiers that reduce the per-invoice rate as you scale. No per-seat fees — there are no users on our side. We don't quote without understanding the use case; a meaningful number requires a 20-minute call.

Is implementation included in pricing?

Implementation for the first environment is in scope of the platform fee. We don't bill hourly — the price you see covers getting you live.

Do you offer a free trial?

We do paid pilots scoped to a single Fusion environment — implementation included. If it doesn't perform against your data, you don't go to production.

Are you SOC 2?

Yes — SOC 2 Type II. Report is available under NDA.

Where does my data live?

Customer data lives in OCI — your tenancy if you want, or a dedicated tenant in our OCI environment if you'd rather not run another OCI subscription. Both options keep data in-region.

Do you train models on our data?

No. We don't train shared models on customer data. Per-customer extraction tuning happens against your data only and stays scoped to your environment.

Multi-tenant or single-tenant?

Customer data is isolated at the storage and compute level. We can operate within your OCI tenancy for customers who require strict single-tenant isolation.

Are you HECVAT-aligned?

Yes — aligned with HECVAT 4.x. We can complete the questionnaire against your specific scope.

We've already built something internally. Why switch?

Most internal builds have done the extraction half well. The chase loop — autonomous, contextual follow-up to suppliers — is where they tend to stop, because it's where AI maturity wasn't there until recently.

If the only piece of your pipeline that still requires people is the chase, that's the conversation worth having.

How do I see a demo?

Book a 30-minute call at cal.com/pypelyne/demo. We don't pitch — we show the pipeline running against a sample environment and answer your specific questions.