The engagement
I led solution definition, architecture alignment, and implementation work for an eKYC and identity platform spanning onboarding applications, backend services, an administrative dashboard, and enterprise identity integrations.
A verified identity has to go somewhere. That hand-off was the centre of the design.
From requirements to a complete journey
The work combined customer workshops, user-journey design, technical documentation, implementation planning, team coordination, and deployment requirements. A central concern was how a verified identity would move through the wider environment, not just how a document or face would be checked.
The solution design connected identity verification, biometric checks, and account provisioning. It also defined the paths for unavailable dependencies, inconsistent data, unsuccessful captures, and cases requiring a reviewer.
Design decisions
Three kinds of user. As requirements evolved, the design distinguished public users, partners, and internal users, with different channels, approval rules, and network boundaries. I worked through these differences with the relevant teams and captured the decisions and remaining questions in the technical documentation.
- Self-service onboarding
- Automatic checks
- Review on exception
- Its own channel
- Approval required
- Rules in refinement
- Separate channel
- Its own network boundary
- Rules in refinement
Read it, don’t retype it. The document-capture flow was designed to extract identity fields automatically, including Arabic and English names and date of birth. Extracted data could then be checked against authoritative identity information and connected to the rest of the verification journey.
- Name (Arabic)
- read from the document
- Name (English)
- read from the document
- Date of birth
- read from the document
- Contact details
- entered by the user
- Confirmation
- the user confirms before submitting
That reduced the need to ask users to re-enter information already present on their identity documents.
Plan for the unhappy path. The design accounted for mismatches, retries, and manual review, rather than treating every captured document as an automatic approval.
- A dependency is unavailable
- Extracted data does not match
- A capture fails
- A case needs a person
- Takes the fallback route
- Retries, then goes to review
- Asks for another capture
- Reaches a reviewer in the dashboard
Sized before it was built. I coordinated infrastructure sizing, environment requirements, network dependencies, and deployment preparation for an on-premises delivery, so the platform’s needs were agreed before installation rather than discovered during it.
The onboarding journey
- 01 · AppCapture the identity document
- 02 · PlatformExtract and cross-check the data
- 03 · PlatformLiveness and face match
- 04 · EnterpriseProvision the account
My scope
- 01
Defined user journeys, business rules, exception paths, and test scenarios with customer and delivery teams.
- 02
Developed architecture and integration specifications across eKYC, identity provisioning, and administrative workflows.
- 03
Led implementation planning across APIs, web and mobile channels, and the dashboard.
- 04
Coordinated infrastructure sizing, environment requirements, network dependencies, and deployment preparation.
- 05
Worked through customer feedback and integration dependencies with the teams responsible for identity, messaging, security monitoring, and operations.
Delivery record
The 2026 project report records delivery of the eKYC API, web application, dashboard, and deployment artifacts. Integration work and end-to-end testing were still active at that point.
Later journey documentation captures further refinement of the different user groups and their access requirements.
