Understanding Content
Lector doesn’t just read documents — it understands them. Text, tables, attachments, handwritten notes, scanned forms.
Your Documents. Automated.
Lector reads, understands and processes incoming documents on its own
– from invoices and applications to contracts.
Whether they arrive by email, scan or fax.
Customers across a wide range of industries rely on lector:
What Is Lector?
Not an OCR Tool.Lector Works on Its Own.
Lector processes incoming documents autonomously — like a digital back office that never sleeps and never calls in sick.
Classifying
Lector recognises what a document is about and who is responsible. Automatically. Without manual rules for every single case.
Extracting
Structured data, ready for line-of-business systems, databases and downstream processes. No more retyping.
Working Autonomously
Lector takes over repetitive processing steps entirely — and escalates only when human judgement is genuinely needed.
How Does Lector Work?
From Inbox to HandoverFully Automated
Lector steps in where people are still sifting, sorting and typing.
Intake
Lector receives documents from every common channel - mailroom, email inbox, scanning line or API.
PDFs
Scans
Letters
Emails
Forms
Attachments
Analysis
Lector reads the whole document: structure, content, intent, context. Not just fields, but meaning.
Processing
Relevant data is extracted, cases are classified, plausibility is checked. All traceable and auditable.
Classification
of cases
of cases
Extraction of
relevant data
relevant data
Validation &
plausibility checks
plausibility checks
Identification &
master-data matching
master-data matching
Handover
Lector hands structured
results straight to your systems, ready for further processing, without anything being re-keyed along the way
Line-of-business systems
APIs
Systems
Mailboxes
500M +
pages processed
and classified
90%+
accuracy in
document classification
<10 sec
processing time
per document
80%+
time saved
during processing
Security & Compliance
Use AI on Your Terms.Secure From Day One.
Lector was built for organisations where data protection and compliance are not optional — they are mandatory. In our European cloud, hybrid, or fully on-premise in your data centre.
GDPR compliant
Encrypted throughout
EU cloud or
on-premise
on-premise
Transparent Processing
Lector explains what it does and why.
No black box.
Auditability
Every processing step is logged. For internal control and external audits.
Roles & Permissions
Granular access control: who may view, edit and approve what.
Where Lector Works
Built forDocument-Heavy Processes.
Document chaos is everywhere. So is Lector.
Frequently Asked Questions
FAQ
That depends on the deployment model. In the Lector Cloud, processing takes place exclusively in European data centres; all data stays within the EU and is transmitted with TLS encryption. With hybrid or on-premise deployment, Lector runs in your own infrastructure.
Then you run Lector on-premise: the platform and the AI models run entirely in your own network, and your documents never leave your data centre. All deployment models – cloud, hybrid and on-premise – are described under Security.
PDFs, scans, letters, emails, forms and attachments – including tables, handwritten notes and scanned forms.
Lector hands structured results straight to your line-of-business systems, APIs, platforms and mailboxes – with no manual re-entry.
Lector combines several specialised models for classification, extraction and plausibility checks — currently Gemini, GPT, Claude, Mistral, InternVL and Qwen, both open and closed source. In the cloud, closed-source models are used exclusively via European servers (Google Vertex AI, Microsoft Azure); on-premise, the models run on your own hardware. Every processing step is logged and traceable.
In the Lector Cloud, GDPR-compliant in Europe; with hybrid or on-premise deployment, on your own servers. Access is controlled through granular roles and permissions and is fully auditable.
REST APIs, connectors for common line-of-business and mailbox systems, plus email and scan intake.
No. The AI models use your data for processing only; with models from the cloud, it is deleted there immediately afterwards.
With Lector, manually reading and retyping documents is a thing of the past. You automate your document processing in a few steps and scale from a hundred to thousands of documents a month. You also benefit continuously from advances in AI models – we build the latest models into the platform for you, so you never deal with API keys, provider changes or rate limits. See for yourself and try it for free.
Single documents, yes. The real question is how you automate the process when hundreds or thousands of documents arrive every month. Lector plugs into your input channels and target systems, sets up new document classes from a description in minutes and makes quality measurable: corrections flow back into the configuration, and in the Evaluation Center you compare settings and models on your real documents. The full comparison is on “What is Lector”.