Tesseract (active)
Self-hosted OCR on your own server, with English and Bangla language packs.
Platform
Passport & travel document OCR
Stop typing passport details by hand. Document Intelligence reads passports, national ID cards, driving licenses and other documents with OCR and proposes values for the form. Staff check and correct the fields, then apply them, in Visa, Hotel, Tour and CRM. Every scan is kept in a history with its status, confidence and both the parsed and the reviewed result.
- 4 modules
- Visa, Hotel, Tour and CRM
- eng + ben
- English and Bangla recognition
- Self-hosted
- Tesseract on your server
- Staff review
- Check before fields apply
Overview
Passport OCR and automatic form filling for travel agencies
Document Intelligence adds passport and travel document OCR to TravelSuite ERP. From a Visa, Hotel, Tour or CRM form, staff click Scan and add the document. The system reads the text, proposes values for the matching fields and waits for a person to check them, so data entry becomes a quick review instead of typing.
It recognises passports, national ID cards, driving licenses and generic documents. The engine available today is Tesseract, running on your own server, with language codes such as eng, ben or eng+ben for English, Bangla or both. Google Vision, Azure AI Vision, AWS Textract and custom providers are listed in settings as coming soon.
You decide how strict the process is. Settings let you switch OCR on or off everywhere, set a confidence threshold, require manual review before fields are applied, turn auto-fill on or off, and choose whether to store the raw OCR text and the OCR logs. The same rules apply in every module that uses scanning.
Every scan is recorded. The dashboard shows total, completed, reviewed and failed scans and a breakdown by module. Scan history lists the module, document type, reference, status, confidence and time, with the parsed and the staff-reviewed result side by side, while OCR logs show the error behind any failed attempt.
How a scan works
From passport photo to filled form, checked by your staff
OCR reads the document and proposes values for the form. Staff compare, correct anything wrong and apply. The parsed and the reviewed result are both kept.
PASSPORTType P · UTO
- Surname
- TRAVELLER
- Given names
- SAMPLE ONE
- Passport no.
- A0000000
- Date of birth
- 01 JAN 1990
- Nationality
- UTOPIAN
- Date of expiry
- 01 JAN 2032
Visa applicantReviewed
- 1TRAVELLER
- 2SAMPLE ONE
- 3
A0O00000A0000000Corrected by staff - 401 JAN 1990
- 501 JAN 2032
Confidence 88%Apply to form
Example data. The passport shown is a fictional specimen.
You stay in control
One settings page decides how OCR behaves
Pick the engine and language, set a confidence threshold and decide what is stored. The same rules apply in Visa, Hotel, Tour and CRM.
Active provider
- Tesseract (self-hosted)Active
- Google VisionComing soon
- Azure AI VisionComing soon
- AWS TextractComing soon
- Custom providerComing soon
engbeneng+ben
Confidence threshold70%
Behavior
- Enable OCRTurn off to hide Scan buttons and block new scans everywhere.
- Manual review requiredStaff review and edit extracted fields before they apply to a form.
- Auto fill enabledFill form fields once a scan is reviewed and applied.
- Store raw OCR textKeep the full text the engine read.
- Store OCR logsKeep a diagnostic log of every scan attempt.
Example settings. Tesseract is the engine available today; the other providers are on the roadmap.
OCR providers
Choose how documents are read
Google Vision
Cloud OCR provider, coming soon.
Azure and AWS Textract
Cloud document AI providers, coming soon.
Custom provider
Connect your own OCR service, coming soon.
Key capabilities
Scan passports, NIDs and driving licenses with OCR and auto-fill Visa, Hotel, Tour and CRM forms after staff review.
12 features included in Document Intelligence (OCR)
See it in a demo- 01OCR for passport, NID, driving license and generic documents
- 02Scan buttons in Visa, Hotel, Tour and CRM forms
- 03Self-hosted Tesseract engine
- 04Language codes eng, ben or eng+ben
- 05Configurable confidence threshold
- 06Manual review before fields are applied
- 07Auto-fill switch for reviewed scans
- 08Option to store raw OCR text and OCR logs
- 09Dashboard with total, completed, reviewed and failed scans
- 10Scan history with parsed vs. reviewed result
- 11OCR logs with error details for troubleshooting
- 12Cloud providers coming soon: Google Vision, Azure AI Vision, AWS Textract
Scan workflow
From document to saved record in six steps
- 01
Scan
Click Scan on a Visa, Hotel, Tour or CRM form and add the document.
- 02
Read
Tesseract extracts the text in the language you set: eng, ben or eng+ben.
- 03
Propose
The parsed values are matched to the form fields with a confidence score.
- 04
Review
Staff compare the values with the document and correct anything wrong.
- 05
Apply
Reviewed fields fill the form when auto-fill is on.
- 06
Record
The scan is saved to history, and any error goes to the OCR logs.
Dashboard, history and logs
Every scan is on record
See OCR use by module, follow each scan from pending to reviewed, and open the logs when a scan fails.
Scans by module
200Total scans
- Visa55%
- CRM20%
- Tour15%
- Hotel10%
200Total scans46Completed148Reviewed6Failed
Scan historyAll modulesAll document typesAll statuses
| Module | Document type | Status | Confidence |
|---|---|---|---|
| Visa | Passport | Reviewed | 94% |
| CRM | NID | Completed | 81% |
| Hotel | Passport | Processing | - |
| Tour | Driving license | Failed | - |
Document types
- Passport
- NID
- Driving license
- Generic
OCR logFailed
Tour · Driving license
Error: check the Tesseract path in settingsExample data.
Built for every team
Less typing for everyone who handles documents
Visa officers
- Passport details proposed for you
- Fewer spelling mistakes
- More files per day
- Corrections kept on record
Front desk
- Scan at the counter
- New CRM contact faster
- Check, then apply
- Less waiting for travellers
Admins
- Self-hosted engine
- Choose what is stored
- Language and threshold settings
- OCR logs for errors
Managers
- Scans by module
- Completed, reviewed and failed counts
- Confidence per scan
- Audit trail of every scan
Why agencies switch
Before and after TravelSuite ERP
| Task | Without TravelSuite | With TravelSuite ERP |
|---|---|---|
| Passport details | Typed by hand from a photo | Read by OCR and proposed for the form |
| Mistakes | Wrong letters in names and numbers | Staff review before fields apply |
| Bangla documents | Typed separately | eng+ben language support |
| Tracking | No record of what was entered | Scan history with parsed and reviewed result |
| Privacy | Files sent to random online tools | OCR runs on your own server |
FAQ
Frequently asked questions
What does Document Intelligence do?
It reads text from travel documents with OCR and proposes values for the matching form fields, so staff review and apply instead of typing.
Which documents can it read?
Passports, national ID cards (NID), driving licenses and a generic type for other documents.
Which modules use it?
Visa, Hotel, Tour and CRM. The dashboard shows how many scans each module made.
Which OCR engine is used?
Tesseract, self-hosted on your server. Google Vision, Azure AI Vision, AWS Textract and custom providers appear in settings as coming soon and are not available yet.
Does it read Bangla?
Yes. Tesseract language codes such as eng, ben or eng+ben let you read English, Bangla or both.
Do staff have to check the result?
If Manual Review Required is on, staff must review and edit the extracted fields before they apply to the form. Auto-fill then fills the fields once the scan is reviewed and applied.
How accurate is it?
Accuracy depends on image quality and the document. Each scan records a confidence score, you set the confidence threshold, and staff review keeps mistakes out of your records.
Can I see past scans and errors?
Yes. Scan history shows every scan with status, confidence and the parsed vs. reviewed result, and OCR logs show the error for any failed scan. Both filter by module, document type and status.
Works with every TravelSuite module
Bookings, customers, payments and accounts stay in one database, so this module shares data with everything else automatically.