Beta Document AI · by QuantumSoft

Answers you can check.

Whatever form your data is in — a PDF, a scan, a phone photo, an Excel file or an email — qDocsAI reads it, understands it and answers with the source.

People ask
Q01

What does qDocsAI do?

Invoices, scans, spreadsheets, emails, contracts: qDocsAI turns them into one set of business data. It knows what each one is, pulls out the figures and parties, links them together and answers with the source.

Invoice_QS-0482.pdfp. 1 / 1

TAX INVOICE

Series QS no. 0482 · 14.10.2026

SupplierExample Distribution Ltd.VAT ID 12345678

CustomerDemo Construct Ltd.VAT ID 87654321

Logistics services, October18,000.00
Freight, two deliveries23,200.00
Handling and storage1800.00

Net totalEUR 12,000.00

VAT 21%EUR 2,520.00

Total dueEUR 14,520.00

Due date: 13.11.2026 · Note: Cash-accounting VAT

What’s the total due?

What’s the total due? The total due is EUR 14,520.00, including EUR 2,520.00 VAT (21%).

Source: Invoice_QS-0482.pdf · p. 1 Confidence 98%
Source: Demo · fictional invoice
Q02

Does it make things up?

It answers from your documents, not from memory, and shows the source under every answer, with a confidence level. You can check any answer in seconds. If one is wrong, you correct it, and the correction is kept.

Live demo · click around

Document assistant

Knowledge base with cited sources

Hi! Ask a question about the indexed documents (invoices, contracts, orders…). Answers include sources from the original files.

Which suppliers do I have the most invoices from?

confidence 100%

The supplier with the highest number of invoices is Atlassian Pty Ltd, with 32 invoices. Here is the breakdown of the top suppliers by invoice count:

  • Atlassian Pty Ltd32
  • Google Cloud EMEA Ltd20
  • Hetzner Online GmbH12
  • Slack Technologies Ltd11
  • CAKE.com AG9
  • Strapi, Inc.8
  • Anysphere, Inc.8
  • Anthropic, PBC4
Sources:124 invoices · type Invoice
Try asking
Source: Interactive demo · Assistant (demo data)
Q03

What if it isn’t sure?

It doesn’t guess. A value it isn’t sure about never reaches your reports. It goes to a review queue, where a colleague checks it against the original document. The decision is kept on record, and qDocsAI learns from it. You decide how sure is sure enough.

Live demo · click around

To validate

What is waiting for your decision: documents the system was not sure about.

INV-NB-2026-0331.pdf

waiting since Oct 23, 3:00 PM

Total due read with low certainty (62%). Check the value or reject the document.

Stage: Field extraction · Certainty: 62%

NCR-2026-0117.pdf

waiting since Oct 23, 3:00 PM

Unknown document type (certainty 95%). Confirm the correct type or reject the document.

Stage: Document type classification · Certainty: 95%

Original document

NON-CONFORMANCE REPORT

NCR No.: NCR-2026-0117Date raised: 21 October 2026

Raised by:
Transmetal Fabricație S.R.L.
Supplier:
Northbridge Steel Distribution Ltd
Purchase order:
PO-2026-0481
Detected at:
Goods-in inspection, Warehouse B
Hollow section SHS 120×120×648 pcs12 rejected
Hot-rolled plate 10×1500×300020 pcs0 rejected

Disposition: return to supplier for replacement.

Check the data before indexing

NCR-2026-0117.pdf

Why we ask you to check

• The document type couldn’t be classified automatically with enough confidence.

Certainty95%

Above the threshold — this is not why the document stopped. · Threshold 75%

How the AI got here

The document is a Quality Management ISO 9001 non-conformance report detailing rejected items from a delivery, which does not match any known document type.

Signals found in the document

  • NON-CONFORMANCE REPORT
  • NCR No.
  • ISO 9001:2015

The AI proposes a type that isn’t in your list yet: Non-conformance report.

Source: Interactive demo · Review queue (demo data)
Q04

Does the format matter?

No. Typed or scanned, a photo or a spreadsheet, an email attachment or an XML export: it’s the same data to qDocsAI. Upload your files or connect Google Drive and OneDrive, and qDocsAI recognises what each one is. We haven’t tested every kind of document yet, and we like a good challenge: bring us yours.

  • Invoices
  • Receipts
  • Orders
  • Contracts
  • Delivery notes
  • Reports
  • Data sheets
  • Emails
  • Digital files (PDF, Word, ODT, RTF)
  • Scans & photos (OCR)
  • Spreadsheets (Excel, CSV)
  • Emails (EML, MSG)
  • Structured exports (XML, HTML, TXT)
Live demo · click around

Documents

All your documents, grouped by type

130 documents2 to validate

Click a document type to open it

Source: Interactive demo · Documents (demo data)
Q05

Can it work on its own?

Yes, on the questions you choose. Routines run a question at the interval you set and put the answer on your home page, next to the state of your documents. Automations can start on an event.

Live demo · click around

qDocsAI Portal

Your routines’ answers and an overview

Novel concepts0Unknown fields awaiting triage
Ingestion errors (24h)0Target under 5% of volume

Your routines’ answers

Expenses in USD, by vendor

Oct 22, 04:24 PM

  • Slack Technologies Ltd1,604.34
  • Atlassian Pty Ltd1,352.2
  • Mosyle Corporation1,098
  • Google Cloud EMEA Ltd846.1
  • Strapi, Inc.588
Suppliers with the most invoices

Oct 21, 02:39 PM

  • Atlassian Pty Ltd32
  • Google Cloud EMEA Ltd20
  • Hetzner Online GmbH12
  • Slack Technologies Ltd11
  • CAKE.com AG9
Source: Interactive demo · Home page (demo data)
Q06

Is it right for my team?

Five quick questions, an honest answer. If you don’t need qDocsAI yet, it will tell you.

Fit check1 of 5

How many documents does your team work with?

Q07

How much time do we spend on documents now?

Move the sliders to your team’s numbers and see the hours it adds up to.

Your numbers
400
6

40hours a month

spent looking things up in documents.

That’s 5 working days, every month.Over a year: 12 full working weeks.

Your numbers, not a promise. qDocsAI is built to cut the looking-up part; how much it cuts depends on your documents, so we measure it with you in the beta.

Q08

Which languages does it work in?

Any language, not just English. qDocsAI was built and tested on Romanian documents, which shows it works natively beyond English, right down to local fields like company tax IDs, cash-accounting VAT and personal ID numbers. If your company receives documents in several languages, they all go into the same system. Ask in your own language and get the answer with its source.

Built on Romanian data: fields it already knows

  • Company tax ID
  • Invoice series and number
  • Cash-accounting VAT
  • Personal ID number
Q09

Which industries is it for?

Any field where work runs on documents and processes: finance, healthcare, construction and more. It started with financial documents, and the same engine works with any data a process produces. If your team follows a process and keeps the paperwork, qDocsAI can work with it.

  • Finance and accounting
  • Healthcare
  • Construction
  • Your process?

For example, in finance and operations

Accounting and invoices

Invoices, receipts, trial balances

“Which invoices use cash-accounting VAT this month?”

Procurement and logistics

Purchase orders, delivery notes, stock reports

“Which orders don’t have a delivery note yet?”

Sales and reporting

Sales reports, CSV registers

“Who are the top 5 customers by value in Q3?”

Contracts

Framework, service and supply contracts

“Which contracts expire in the next 60 days?”

Across documents

Delivery notes, goods receipts and invoices, read together

“Which deliveries were invoiced but never received?”

Q10

Who on my team sees what?

Each role has its own dashboard and access only to the document types it needs. Each company’s data is kept separate, and sign-in works with email or a Google account.

  • Administrator
  • Manager
  • Accountant
  • Analyst
  • Operator
  • Employee
Q11

How much does it cost?

There’s no public price yet. qDocsAI is in beta, and we’re agreeing terms individually with each beta partner. Our aim is to keep the price as low as possible, so it’s within reach for companies of any size.

Q12

How do I try it?

We work with companies of every size, from small teams to large organisations, and our beta partners’ feedback helps shape the product. Tell us about your team and we’ll get back to you with the details.

Request beta access