Live · Production-ready

Universal Document
Intelligence.

One platform for every document: read, compare, automate, monitor, with a person checking only the cases the model flags as uncertain.

< 5s
avg. doc processing
99.2%
field-level accuracy
EU
hosted, GDPR
Day 1
production go-live

Live preview

Click a package. Watch the product change.

Every package below switches the live preview to its real configuration. No mockups, no cropped screenshots, this is the actual product.

app.koeniq.de/udx/hitl
Standard · HITL-UI
Mensch in der Schleife abKonfidenz < 95%
2 auto0 in Schleife
Queue · 3
Rechnung_2024-03.pdf
Vertrag_Mueller.pdf
Eingangspost_42.pdf
+ 142 ältere
RECHNUNG
Rechnung Nr. 2024-0314
Alpha GmbH · Kundennr. 4821

Lieferant: Alpha GmbH, Maximilianstr. 12, 80539 München

Steuernr.: 143/815/08150 · USt-IdNr.: DE 264 847 276

Rechnungsdatum: 14.03.2024 · Fällig: 28.03.2024

Rechnungsnr.: 2024-0314 · Auftrag: AO-2024-088

─────────────────────────────────────────────

Pos. 1 Wartung Server Q1/2024 1.200,00 €

Pos. 2 Monitoring & Alerting (3 Mon.) 840,00 €

Pos. 3 Notfall-Support Pauschale 346,55 €

─────────────────────────────────────────────

Netto: 2.386,55 € · MwSt 19%: 453,45 €

Gesamt: € 2.840,00

Zahlbar auf IBAN DE89 3704 0044 0532 0130 00.

BIC: COBADEFFXXX · Commerzbank AG

AUTO-FREIGEGEBEN
Extrahierte Entitäten
4 Felder · 4 ok
Lieferant99%
Alpha GmbH
Betrag98%
€ 2.840,00
Datum97%
14.03.2024
IBAN99%
DE89 …0130 00

Pakete

Sofort einsatzbereit. Modular erweiterbar.

Klicken für Live-Vorschau
Sofort einsatzbereitVorschau aktiv
Standard

Alles, was du brauchst, um ab Tag 1 live zu gehen. Die KI liest deine Dokumente, extrahiert Entitäten automatisch und leitet alles, wo sie sich nicht sicher ist, an einen menschlichen Prüfer weiter.

Entity-Detection (KI)Entity-ConfigurationudX-EngineHITL-UI
Buchhaltung & Controlling
Finanz-Paket

Alles aus Standard, vorkonfiguriert für Rechnungen, Belege und ERP-Buchung. Duplikate und Anomalien werden automatisch erkannt und geblockt, bevor sie ins System gelangen.

Entity-Detection (KI)Entity-ConfigurationudX-EngineHITL-UI
+Anomalie-Erkennung+Widerspruchs-Erkennung+OneDrive-Integration
Recht & Compliance
Law-Paket

Alles aus Standard, vorkonfiguriert für Vertragssprüfung und Compliance. Klauseln werden extrahiert, Widersprüche zwischen Dokumenten erkannt, und dein Team kann direkt mit dem Archiv chatten.

Entity-Detection (KI)Entity-ConfigurationudX-EngineHITL-UI
+Klausel-Extraktion+Widerspruchs-Erkennung+Cross-Document-Linking+Chat w/ Docs

Erweiterungs-Module, einzeln dazubuchbar

+ OneDrive
+ Widerspruch
+ Anomalie
+ Cross-Doc
💬+ Chat
+ Foundry

Wie funktioniert es

Von der Datei zur Entscheidung.

Deine DokumentePDF · Mail · Scan · Excel
StandardudXDetection · Config · Engine · HITL
Strukturierte DatenFreigabe · Weiterleitung · Buchung
Hover auf einen Schritt für mehr Details
Strategiegespräch buchenKostenlos · 45 min · du gehst mit einer Roadmap raus

What it does

Four things udX does well.

Extraction

Every field, every layout, every format: PDF, scan, email, Word, image. Extracted into structured data with a confidence score per field.

Comparison

Two contracts, ten policies, a notice against the declaration. udX surfaces every meaningful diff and flags what matters.

Automation

Triggered on receipt: extract, validate, route, post. Humans only on the rows the model isn't sure about.

Monitoring

Deadlines, anomalies, missing fields, expiring contracts. udX watches and alerts before things slip.

How it works

From inbox to system of record.

Four stages. Each one inspectable, each one swappable, each one already production-tested.

01

Ingest

Documents arrive via email, upload, OneDrive, S3, or your existing inbox. udX accepts every common format with no preprocessing.

02

Understand

OCR and entity detection extract every relevant field. Domain-tuned models score confidence per field, so the engine knows when it doesn't know.

03

Verify

Anything below threshold lands in the HITL queue: a human approves or corrects it in seconds, and the model learns from that correction.

04

Deliver

Structured data flows into your ERP, CRM, DMS, or storage of choice: REST, CSV, DATEV, Lexware, SAP, whatever you run.

Integrations

Plays well with what you already run.

DATEVLexwareSAPOneDriveSharePointS3REST APIWebhookCSV / XMLEmail (IMAP)Microsoft 365PostgreSQL

Don't see your stack? udX exposes a clean REST API and webhook layer, so anything you can call, you can connect.

Why it works

Built for production, not for screenshots.

Human-in-the-loop, by design

udX never silently approves what it isn't sure about. Every uncertain field is flagged for review, and every correction trains the next decision.

EU hosting, GDPR-first

Hosted in Germany. No data leaves the EU. Encryption at rest and in transit. Full audit trail per document.

API-first, no lock-in

Every action udX takes is reachable via API. Export your data and your processed results in one call, any time.

The udX family

Two products build directly on the engine.

Every document analysed by udX is immediately usable in both: same entity model, same hosting, same human-in-the-loop principle.

Pricing

Three packages. Or build your own.

See full pricing →
Standard
from €8.900/ mo

Production-ready out of the box. AI extraction, HITL, standard configuration.

Finance
from €13.900/ mo

Pre-tuned for accounting. Anomaly and contradiction detection.

Law
from €17.900/ mo

Pre-tuned for legal. Cross-document linking, chat with docs.

Ready to ship it?

45 minutes. We look at three of your real documents. You walk away knowing exactly what udX would do with them, and what it would cost.

Book your call

Frequently asked questions about udX

What is udX?

udX (Universal Document Intelligence) is Koeniq’s platform that automatically reads, understands and processes any document. Four modules cover extraction, comparison, automation and monitoring, with no templates and no OCR setup. Processing in under 5 seconds at 99.2% field accuracy.

How is udX different from traditional OCR?

OCR only recognises characters and needs a fixed template per document type. udX understands the content regardless of layout: it extracts the right fields even from unknown formats, maps them correctly and detects relationships, without any predefined templates.

How is udX different from RPA?

RPA replays rigid, predefined click paths and breaks the moment a UI or document changes. udX works content-based and stays robust against layout changes, and hands uncertain cases to a human instead of running on incorrectly.

Which documents can udX process?

Invoices, receipts, contracts, insurance claims, tax notices, annual reports, doctor’s letters and application forms, among others. Typical fields are legal, tax and finance, insurance and healthcare, including integration with systems such as DATEV or Lexware.

Where is udX data hosted?

Exclusively in the EU and GDPR-compliant. For uncertain results, udX flags the case for manual review (human-in-the-loop), so no incorrect data enters your systems unnoticed.