Persons
What is the Person System?#
The person system is a centralized solution for automatic collection, storage and management of user profiles based on the results of completed technologies.
Person System Architecture#
Person — Central Entity#
- Main profile of an individual in the system
- Organization binding — each person belongs to a specific organization
- Verification status (
is_verified
) — identity confirmation - Flow session connection — tracking all verifications the person participated in
- Uniqueness within the organization is ensured through the alias system
- Automatic creation of system ID when creating a person
- Recovery support — ability to restore a "deleted" person
Aliases — Identification System#
- Multiple identifiers of one person for various search methods
- Alias types:
SYSTEM_ID
— internal UUID, created automatically, immutablePHONE
— phone number from eDocument resultsPERSONAL_NUMBER
— IIN/personal number from documentsDOCUMENT_NUMBER
— passport or identity card numberCUSTOM
— custom identifiers from metadata
- Uniqueness guarantee within the organization
- System alias protection — cannot delete/modify SYSTEM_ID
- Automatic extraction from document verification results
- Conflict checking when adding new aliases
- Search and deduplication — main mechanism for preventing duplicates
Photos — Multimedia Profile#
- Various sources of person images
- Photo types by priority:
DIGITAL
(priority 1) — digital photo from document chipSCAN
(priority 2) — portrait from scanned documentLIVE
(priority 3) — best frame from liveness verificationOTHER
(priority 4) — manually uploaded photos
- Default photo (
is_default
) — automatic selection by type priority - Default photo protection — cannot delete the only default photo
- Automatic management — when adding a photo with highest priority, it becomes default
Actions — Audit and History#
- Complete logging of all operations with the person
- Action types:
- Person creation
- Verification status change
- Alias operations
- Photo operations
- Person deletion
- Action sources:
FLOW
— automatic actions from flow sessionsAPI
— actions through Backend APICABINET
— actions through cabinet web interface
- User binding — who performed the action (if applicable)
- Timestamps — when the action was performed
- Immutability — actions cannot be changed or deleted
Person Configuration — Behavior Settings within Flow#
- Flow technology binding — settings for specific verification flow
- Collection management:
- Enable / Disable automatic person collection
- Force verification of new persons
- Skip technologies for known persons
- Use person photo for Face2Face verifications
Integrations and Dependencies#
- Subscription system — control of person creation limits
- Flow sessions — source of automatic data
- Verification technologies — EDocument, Liveness, Document Recognition
- Organizations — data isolation between clients
- Users — binding actions to specific operators
Automatic Data Collection and Processing#
Extraction from EDocument:
- Automatic recognition of personal data from electronic document retrieval results
- Extraction of IIN, document number, phone number
- Saving digital photo (DIGITAL) from electronic document
Extraction from Liveness:
- Saving the best frame (Live)
Extraction from Document Recognition:
- Extraction of personal number, document number
- Saving portrait photos (SCAN) from document
Ways to Work with the Person System#
Backend API — Integration with External Systems#
Purpose: Programmatic integration with external systems for creating and managing persons through REST API.
Detailed description can be found here
Typical Use Cases#
- Data migration from existing systems
- CRM integration for creating client profiles
- Batch loading of persons from external sources
- Synchronization with third-party databases
Cabinet (Web Interface) — Manual Management#
Purpose: Web interface for administrators and operators for manual person management through browser.
Available Operations:#
- View person list
- Detailed person view
- Person creation
- Photo management
- Alias management
Web Interface Features#
- File upload — direct image upload through form
- Photo preview — ability to view all images
- Advanced filtering — multiple search criteria
- User audit — all actions are linked to user
Typical Use Cases#
- Manual correction of person data
- Administrative management — bulk operations
- Incident investigation — analysis of action history
- Quality control — checking automatically created persons
- Operator training — convenient interface for new employees
Flow (Automatic Collection)#
Purpose: Automatic creation and updating of persons based on completed flow session results.
Operating Mechanism#
- Asynchronous processing — tasks are executed in background
- Launch trigger — automatic start after flow session completion
- Configurability — behavior is configured through Flow technology configurations
Automatic Collection Algorithm#
1. Data extraction from results
- EDocument results → IIN, document number, phone, DIGITAL photo
- Document Recognition → personal number, document number, SCAN photo
- Liveness results → LIVE photo (best frame)
- Session metadata → CUSTOM alias (if specified in flow session metadata)
2. Search for existing persons
- Search persons by all extracted aliases
- If one person found → update
- If none found → create new person
3. Person creation/update
- Filter new aliases (exclude existing ones)
- Filter new photos by type
- Set default photo by type priority
- Determine verification status (Face2Face result)
Configuration through Person Config#
Detailed behavior:
Person collection disabled
- Persons are not created automatically
- Flow sessions complete without data collection
- Manual creation through API/Cabinet remains available
Force person verification
- All new persons are created with
is_verified = True
- Face2Face verification result is ignored
Skip technologies for existing results
- Search person by flow session metadata
- Copy results from previous technologies (EDocument)
- Skip re-execution of already completed verifications
- Speed up repeat verifications for known persons
Use for Face2Face
- Use person photos for Face2Face comparison
- Automatic selection of best available photo
- Improve biometric comparison accuracy
Photo Processing Priorities#
- DIGITAL (priority 1) → default photo
- SCAN (priority 2) → if no DIGITAL
- LIVE (priority 3) → if no DIGITAL and SCAN
- OTHER (priority 4) → lowest priority
Automatic Processing Scenarios#
Scenario 1: New user
Flow: EDocument + Liveness
Result:
- Create person with IIN and document number
- Two photos: DIGITAL (default) + LIVE
- Verification default false
Scenario 2: Repeat verification
Flow: EDocument + Liveness + Face2Face
Result:
- Update existing person
- Add new LIVE photo (if type didn't exist)
- Verification = true (successful Face2Face)
Scenario 3: Fast verification - Skip technologies for existing results
Person found by alias in metadata
Result:
- Copy EDocument result from previous session
- Skip re-execution of EDocument
- Execute only new technologies
Comparison of Working Methods#
Characteristic | Backend API | Cabinet UI | Flow |
---|---|---|---|
Initiator | External system | Human operator | System (automatically) |
Operation type | Programmatic | Interactive | Automatic |
Photo upload | Base64 | File through form | Extraction from results |
Authentication | API key | Web session | - |