Skip to content

EgyIDOCR

Document intelligence for Egyptian national IDs and passports, combining several OCR providers with AI post-processing.

Sector
Enterprise
Stage
Live product
Client
ITechPro product

The challenge

Organisations that register people — hotels, employers, service providers — re-key ID and passport details by hand, which is slow and error-prone.

The solution

  • Extraction from Egyptian national IDs and passports.
  • Several OCR providers combined, with AI post-processing to correct and structure the result.
  • A review step for low-confidence reads before data reaches a system of record.

Capture

Images of IDs and passports from web or mobile.

Extraction

Multi-provider OCR with AI post-processing.

Review & export

Human review below a threshold, then structured output.

Technology architecture

  1. Access

    • Web capture
    • Mobile capture
    • API
  2. Services

    • OCR providers
    • AI post-processing
    • Review queue
  3. Platform

    • ASP.NET Core
    • Next.js
    • Audit log
Reference architecture showing how a system of this kind is structured. Not the deployed topology.

Implementation approach

How we structure delivery for a system of this kind.

  1. 01

    Use more than one OCR engine and compare results.

  2. 02

    Correct and structure the output with an AI model.

  3. 03

    Send uncertain reads to a person instead of guessing.

Outcomes

We publish measured results only once the client has verified them. Ask us about this project and we will share what we can.

Reviewer note: No measured outcomes have been approved for publication yet. Add verified figures here.

Source: Information supplied by ITechPro.