Step 1
1. We identify the country and document type
The neural network classifies the document from a photo or a scan, corrects the perspective and removes glare and noise.
Document recognition powered by AI-OCR
In a second, NeuroVision extracts data from passports, driving licences, ID cards and forms of 195+ countries and territories. The result is ready JSON fields through the REST API or an SDK.
Data is stored in Russia Russian Software Registry No. 19972 Compliance with 152-FZ
What NeuroVision AI-OCR recognises
NeuroVision AI-OCR supports more than 16,000 document types from 195+ countries and territories. The model identifies what it is looking at by itself, so the user does not have to choose the document type manually.
- Passport recognition
Russian internal and international passports
The photo page, the registration page and the MRZ — every version of the form.
-
Driving licences, ID cards, residence and temporary residence permits
Russian and foreign licences, national ID cards, residence permits and temporary residence permits. Front and back sides.
- The full list of documents
Documents of 195+ countries and territories
Passports and ID cards of the CIS, Europe, Asia and the Middle East, including Latin, Cyrillic, Arabic script and hieroglyphs.
-
MRZ, barcodes and QR codes
We read the machine-readable zone, PDF417 on licences and ID cards, and QR codes. Data from MRZ is cross-checked against the visual zone.
-
Forms, contracts, invoices and banking documents
Template forms, applications, statements, utility bills, social security and tax numbers. We extract fields using the customer's template.
- Handwriting recognition
Handwritten text
Forms and applications filled in by hand in block letters.
How document recognition works
The whole cycle takes about a second; a person steps in only when the model's confidence is low.
Step 2
2. We extract the fields
The model locates every field and reads it taking the language and typeface into account. Dates, numbers and codes are brought to a single format.
Step 3
3. We check the MRZ, check digits and signs of forgery
We cross-check the MRZ against the visual zone, calculate checksums, verify the expiry date and look for traces of editing. More in the section customer verification (KYC).
Step 4
4. We return the result in JSON
Fields, the model's confidence for every field, a crop of the document photo and the status of the checks. The data goes straight into your CRM, core banking system or form.
More than OCR: intelligent document processing (IDP)
Ordinary OCR turns an image into text but does not know which part is the surname and which is the document number. NeuroVision AI-OCR works as an IDP system: it understands the structure of the document, sorts the data into fields and validates it.
| Criterion | Ordinary OCR | AI-OCR NeuroVision |
|---|---|---|
| Result | Continuous text | Fields in JSON |
| Document type | Chosen by the user | Detected automatically |
| Phone photo, glare, tilt | Errors | Alignment and clean-up |
| MRZ and codes | Not read, or read separately | Read and cross-checked against the visual zone |
| Forgery check | No | Check digits, expiry dates, traces of editing |
| Integration | Manual transfer | REST API, SDK, webhooks |
Where document recognition is used
Recognition fits into onboarding, paperwork and record-keeping — everywhere employees or customers currently copy data from documents by hand.
-
Banks and MFIs
- Remote issuing of cards and loans
- The passport is recognised inside the form — the customer enters nothing by hand
-
Insurance
- Issuing policies online and settling claims
- Licence, vehicle registration, passport
-
Marketplaces and e-commerce
- Verification of sellers and buyers
- Age verification for 18+ sales
-
Car sharing and rental
- Licence checks at sign-up in a minute
- With no office visit
-
Healthcare and front desks
- Registering a patient by passport and health insurance policy
-
Logistics, HR and personnel records
- Onboarding couriers and employees, including foreign nationals
- Passport, work patent, residence permit
Why a business needs automatic document recognition
- 01
Higher onboarding conversion
The customer photographs the document instead of typing in 15–20 fields.
- 02
Fewer errors
No operator typos; the data is cross-checked against the MRZ.
- 03
Faster processing
A second instead of 3–5 minutes of manual entry.
- 04
Less fraud
Forged and edited documents are filtered out before the decision.
Results of deploying document recognition
A case study from a large Russian bank
- +15% to onboarding conversion
- down to 0.01% fraud rate
- ×10 faster document processing
How much an OCR service costs
The price depends on the volume of checks and the integration method. You pay only for recognised documents, with no subscription fee for idle time. The first month of testing is free.
- Calculate the cost
Cloud (API/SDK)
Payment per document
Volume discounts: the more documents per month, the lower the price per recognition.
- Calculate the cost
Enterprise
Custom terms
A dedicated manager, an SLA and 24/7 support.
- Calculate the cost
On-premise
Licence
Deployment inside your perimeter: the data never leaves the company's infrastructure.
Why NeuroVision
Document recognition is part of the customer verification platform: the same models, infrastructure and security requirements.
NeuroVision — an expert in computer vision and AI
AI-OCR works as part of customer verification (KYC), trusted by the largest companies in fintech, e-commerce and transport services.
We develop identification algorithms that automate checks, detect fraud and help meet regulatory requirements.
A Russian developer
A Skolkovo resident, software in the Russian Software Registry (No. 19972), Rospatent patent No. 2023661508.
OCR API and SDK: live in a single day
A ready-made capture interface, a REST API or deployment inside your perimeter — choose the integration that suits your architecture.
- Fields in JSON
- Confidence for every field
- Sandbox and documentation
- Webhooks
Terms of connection
A free one-month test, a dedicated manager, 24/7 support, payment only for checks performed.
Deployment options for the document recognition service
Connect document recognition in the cloud in a single day — or deploy AI-OCR inside your own isolated perimeter in line with regulatory requirements.
A quick start with no infrastructure
payment for recognised documents, scaling to the load, updates without your involvement.
Data is processed in data centres in Russia.
Full isolation
the service is deployed on your servers; documents and data never leave the protected perimeter.
Integration with internal systems, compliance with 152-FZ and 242-FZ.
Easy integration into your stack in an hour
A REST API with detailed documentation and an SDK with a ready-made capture interface — start recognising documents on the day you connect
Licences and certificates
The project «Development of a multimodal system using artificial intelligence "Neuromed" for managing the diagnostic processes of a clinic on the basis of the "Neurovision" computer vision platform» is supported by the Innovation Promotion Foundation within the federal project «Domestic Solutions»
The Neurovision computer vision platform is registered with Rospatent (certificate No. 2023661508) and included in the Russian Software Registry (entry No. 19972 of 13.11.2023)
The right to use the Platform is granted under a licence agreement concluded by Neurovision with its customers
Try document recognition for free
Leave a request — we will give you a month of test access to the API and SDK and help with the integration.
Can't put your request into words? Book a call and our team will contact you at a convenient time
Book a call
The company is included in the register of IT companies and the register of personal data operators of the Russian Ministry of Digital Development.
Address: Moscow, Presnenskaya Embankment 12 — Moscow City business centre, Federation Tower. Nearest metro stations: Delovoy Tsentr and Mezhdunarodnaya.
Read also
Materials on document recognition, the MRZ and AI-OCR accuracy — in our blog.
385 13
KYC Document Verification
How KYC Systems Detect Forged Documents
354 17
KYC Document Verification
Passport Verification Online: How Automatic Verification Works in KYC
363 21
KYC Document Verification
Why users retake documents: how to reduce the number of repeat attempts during a KYC check
510 13
KYC Document Verification
Passport machine-readable zone verification: how the MRZ helps detect forgeries in KYC
Frequent questions about document recognition
Answers about document types, accuracy, speed, data storage and integration
Russian internal and international passports, driving licences, ID cards, residence and temporary residence permits, and more than 16,000 document types from 195+ countries and territories, plus MRZ, barcodes, QR codes and template forms.
Field recognition accuracy is 99.85%. The service returns a confidence score for every field, and doubtful cases can be sent for manual review.
About a second, including classification, field extraction and the MRZ check.
Yes. The SDK prompts the user on how to capture the document, and the model aligns the image and removes glare.
Ordinary OCR returns continuous text. AI-OCR understands the structure of the document, sorts the data into fields and validates it.
Data is processed on servers in Russia. The service complies with 152-FZ, and with on-premise deployment the data never leaves your perimeter.
Yes, there is an on-premise option: the service is deployed inside your perimeter.
You pay for recognised documents, and the first month is a free test. We will calculate the exact price for your volume.
