BrandKwikID Documentation

Architecture

Platform design, capability boundaries, and deployment topology for Kwik Vault

Architecture and Technical Design

Kwik Vault ships as a unified platform with clear capability boundaries inside a single deployable core. Supporting data and infrastructure services run alongside the application tier.

Overview

Each bank receives a dedicated Deployment Instance. The platform ingests documents from integrated source systems (KYC flows, document services, branch uploads), stores file content in object storage, indexes metadata in a search index, routes approvals through a workflow engine, and records an append-only audit chain in a relational store.

High-Level Architecture

Pan, scroll, or pinch to zoom · use controls for fit and zoom

Platform Capabilities

CapabilityRole
IngestAccept uploads; validate metadata; store blobs; create versions
MetadataDocument schema definition and validation
SearchFederated customer and document discovery
WorkflowApproval state machine and durable lifecycle orchestration
AuditAppend-only audit entries with integrity chain
TenancyOrganization, department, and branch scope
Access ControlPermission checks and department-scoped RBAC
IdentityUser provisioning and integrator credential management
AIPluggable natural-language discovery over governed corpus

Metadata and access control are shared foundations consumed by other capabilities; they do not expose a separate public interface.

Infrastructure Layer

ComponentRole
Web PortalUpload, documents, approvals, audit, and configuration UI
Core Platform APIREST API surface for portal and integrators
Workflow ProcessorRuns approval and lifecycle workflows
Relational StoreDocuments, tenants, workflow state, and audit chain
Object StorageVersioned document blob store
Search IndexMetadata and full-text retrieval
Workflow EngineDurable orchestration for multi-step approvals
Identity ServiceSSO federation and machine-to-machine authentication

Tenancy Model

  • One Organization per Deployment Instance.
  • One Operational Tenant plus zero or more Restricted Tenants for compliance, audit, and treasury isolation.
  • Federated Search returns only documents in tenants the user is authorized to access.
  • Cross-tenant access is always logged in the Audit Trail.

Document Identity

ConceptDescription
DMS IDSystem UUID for each Document; shown in Documents and Audit tables
Business Keysource_tag + external_id (e.g. vkyc:session-123)
VersionRe-ingest with the same Business Key creates a new version
Current VersionLatest approved version; pending amendments do not replace it

Logical Folder Structure

Documents are stored under a deterministic logical hierarchy (not a user-managed folder tree):

{organization}/{department}/{document_category}/{customer_id}/{business_key}/v{version}
SegmentMaps toExample
Customer IDCIF, PAN, or account number from customer_refs12345678
Document typedocument_category ingest fieldkyc_vkyc
Parent IDexternal_id or custom_attributes.parent_id for child documentsloan-app-456

Re-ingesting with the same Business Key creates a new version on the same DMS ID without overwriting prior blobs. See Storage and Lifecycle for ILM tiering and Operations Guide for department onboarding.

Version Approval State Machine

Benefits

  • Scalable design: Core capabilities can scale independently as document volume grows.
  • Proven infrastructure patterns: Object storage, search indexing, workflow orchestration, and identity federation use industry-standard roles.
  • Bank-grade isolation: Dedicated instance per bank prevents data commingling across institutions.

Next Steps