Documents & drawings
Extract information from emails; read and check e-invoices.
HEROextractor
INVOICETECHNICAL
DRAWINGEMAILInvoice numberINV-2048
Total amountEUR 1,250.00
SupplierS-1042
HEROextractor · Extract document data with AI
Review instead of retyping.
HEROextractor extracts item numbers, measurements, invoice amounts and tables from your documents, images and text, including electronic invoices. If desired, suitable categories such as cost centres or quality grades are suggested.
Print & handwriting
Recognise content and extract relevant fields.
Structured data
Fields, tables and matrices for your workflows.
Turn content intoinformation.
See how a document becomes usable data. Explore the process with fictional examples, then try your own document below.
TechSolutions GmbHRosenweg 5 · 10115 Berlin
Invoice
SAMPLE LTD Industriestraße 12 80953 Munich
Invoice date: 14 March 2024
Supply of technical components
| Description | Quantity | Unit price | Amount |
|---|---|---|---|
| Sensor module | 5 | EUR 160.00 | EUR 800.00 |
| Assembly kit | 1 | EUR 250.42 | EUR 250.42 |
Net amountEUR 1,050.42
VAT 19%EUR 199.58
Payment due within 14 days.
HEROextractor
Extracted data
| Data field | Example value |
|---|---|
| Invoice number | INV-2048 |
| Total amount | EUR 1,250.00 |
| Supplier ID | S-1042 |
Select a field on the left to highlight its matching value.
Give the input a context.
Classification distinguishes document types. Recognition of print and handwriting makes content available for further processing.
DOCUMENTS COMPARED.
HEROextractor compared
Loading results …
Loading results …
TRY IT YOURSELF
The best demo is your own document.
Upload your file, choose the information you need and check the results against the original. Free, with no sign-up.
100% processed in the EUOperated by DOCSHEROThe document demo is loading. Please enable JavaScript to use it.
Start right away.Improve with purpose.
Default templates work out of the box. For unusual documents, extra fields or greater accuracy, extend HEROextractor and train it with your own examples. The Training & Testing Suite helps you assess progress before going live.
Train
Use your own sample documents to tailor recognition to your needs.
Put everything in the right category.By your rules.
After extraction, you can also categorise the data – for example, by cost centre or quality grade. You define the criteria, and AI suggests suitable categories.
Extracted supplier
NameBeispiel GmbH
IDS-1042
Master data
Matched supplier
NameBeispiel GmbH
IDS-1042
AI categories & rules
For example: office supplies → cost centre 100 “Administration”. You define the criteria, and AI suggests the allocation. Business rules validate the extracted values.
Matching Suite
Match extracted information against your master data, such as supplier or customer numbers. This connects the results to your existing systems.
Original and results.
Side by side.
Check extracted values, tables and matrices against the original and correct them where needed. Stay in control of the data you use in your next steps.
Your solution in three steps.
Understand your needs
Define inputs, required fields and target systems together.
Validate with samples
Use representative documents, test rules and agree measurable quality goals.
Integrate into your workflows
Configure data transfer, validation and operations for your business processes.
Could this workfor your team?
Show us the documents you process and the information you need. Together, we’ll work out how HEROextractor fits into your workflows.
Discuss your use caseYour examples. Your workflows.
Which documents can we extract data from automatically?
Examples include invoices, delivery notes, contracts, technical drawings and emails. HEROextractor processes digital documents, scans, images and text, including printed and handwritten content. Representative samples are the best way to assess how well it works with your documents.
Can we choose which data fields to extract?
Yes. You decide which information you need and can add or adjust fields, descriptions and table columns. Alongside individual values, HEROextractor supports tables and matrices, for example for invoice line items, quantities or measurements.
Do we need to train HEROextractor before getting started?
For supported standard document types, you can start with the existing templates. If information is missing or your particular documents need better recognition, you can add fields and rules and train models using your examples. This lets you improve the results that matter to your daily work.
What happens if a value is read incorrectly?
You can review and correct extracted values alongside the original document. Rules and comparisons with your master data help identify implausible values. Together, we define the checks your process needs before the data is passed on.
Can HEROextractor work with our existing software?
Extracted information is available as structured data for further processing. REST interfaces can connect applications and databases and compare values with existing master data. We work with you to establish which connection your software supports and where each piece of data should go.
Where are our documents processed?
Processing is carried out entirely by DOCSHERO. All services run 100% in the EU on servers rented by DOCSHERO. No external apps or AI providers are involved in processing.
Can we test HEROextractor with our own documents?
Yes, without signing up, in the “Try it yourself” section on this page. Upload a PDF, image, text file, e-invoice or XML/JSON file, select the document type and adjust the fields if needed. Then review the results alongside the original. You can also request a personal demo to discuss your specific workflows.
Do the online demo limits also apply in production?
No. The free online demo allows files up to 1 MB, up to 10 pages and 16 documents per person per hour. This protects the public demo from overload. These limits do not apply in production, where we tailor processing to your document volumes and workflows.
Can HEROextractor also suggest cost centres?
Yes. You provide your cost centres and describe which expenses belong to each. Based on the extracted information, AI suggests an allocation – for example, office supplies to cost centre 100 “Administration” or machine spare parts to 200 “Production”. The cost centre does not need to appear in the document: it is suggested using your criteria and can be reviewed before further processing.
Are e-invoices supported and checked?
Yes. HEROextractor supports e-invoices such as XRechnung and ZUGFeRD/Factur-X. It reads the structured invoice data and checks the information against your business rules. In the verifier, you can review the results alongside the document and correct them before further processing.
Loading form…