Product · Identity and trust infrastructure
Daju Verify is identity and trust infrastructure for organisations that have to be sure.
Verification is rarely one check. It is a sequence that differs by country, by risk level and by what the customer can actually provide — and every step of it has to be explainable afterwards. Daju Verify orchestrates that sequence, records the evidence, and gives your reviewers a case to decide rather than a raw score.
Case review queue
Evidence
Risk signals
Decision trail
Positioning is drafted from the product brief and is awaiting confirmation. Daju Verify makes no claim of government approval, accreditation, or guaranteed fraud prevention.
The problem
Organisations operating across several countries end up with a different verification process in each, built on whichever provider was available at the time. The result is inconsistent decisions, evidence scattered across systems, no single audit trail, and a change of provider that costs a re-integration. Meanwhile customers abandon onboarding because the process asks for documents they do not have in a format they cannot produce.
The product insight
The durable asset is the orchestration and the evidence, not the individual data source. If verification is modelled as a configurable sequence of steps with recorded outcomes — and every provider sits behind an interface — then adding a country, changing a supplier or tightening a risk rule becomes a configuration decision rather than an engineering programme, and the audit trail stays continuous across all of it.
Who it is for
- Financial services and fintech
- Firms with customer due diligence obligations that need consistent, auditable onboarding across several markets.
- Marketplaces and platforms
- Platforms verifying sellers, providers, drivers or professionals before allowing them to transact with users.
- Public sector and regulated services
- Organisations confirming eligibility or entitlement, where the decision has to be explainable and appealable.
- Employers and staffing
- Organisations verifying identity, right to work and professional credentials as part of onboarding.
Value
What it gives you
- One integration; verification sequences configured per country, product and risk level.
- Document, biometric and data-source checks orchestrated as steps, not hard-coded flows.
- Consent captured and recorded explicitly at the point it is given.
- A complete, immutable audit trail for every decision, including the human overrides.
- Case review that gives an assessor the evidence and the reasoning, not just a number.
Capabilities
What exists today, and what does not
Split by status deliberately. A roadmap item presented as a shipped feature is the most common way software companies lose trust.
Live · 1
Consent and audit trail
Consent captured at the point of collection with its scope, version and timestamp. Every check, decision and override is recorded in an append-only trail.
In beta · 7
Verification orchestration
Configurable sequences of verification steps by country, product and risk level, with conditional branching on the outcome of earlier steps.
Document verification
Identity document capture, authenticity assessment and data extraction, with per-field confidence and a review queue for anything below threshold.
Biometric verification
Selfie capture with liveness detection and comparison against the document portrait. Biometric processing is treated as special-category data throughout, with explicit consent and defined retention.
Business verification
Company registry lookups, ownership and control structure, and verification of the individuals behind a business.
Fraud and risk signals
Device, behavioural and data-consistency signals combined into a risk assessment that is explainable at the component level rather than delivered as an opaque score.
Case review
A reviewer workspace showing the evidence, the signals and the reasoning, with decisions, notes and overrides recorded against the reviewer.
Developer APIs and webhooks
REST APIs with signed, idempotent, replay-resistant webhooks, retry and dead-letter handling, and a sandbox with deterministic test identities.
Planned · 1
Country-specific data sources
Per-market data-source integrations behind a common interface. Availability differs by country and is stated per market rather than implied globally.
Journeys
How people actually move through it
Completing verification without abandoning
- 01Start verification from the organisation’s own onboarding flow.
- 02See what will be checked, why, and what happens to the data — before providing anything.
- 03Give explicit consent, recorded with its scope and version.
- 04Capture a document and, where required, a selfie with liveness.
- 05Receive a clear outcome: verified, needs review, or a specific action to take.
- 06Where review is needed, be told what is happening and roughly how long it will take.
Deciding a case with the evidence in front of you
- 01Open the review queue, prioritised by risk and age.
- 02See the full evidence: documents, extracted data, signals and check outcomes.
- 03See which checks passed, which failed, and specifically why.
- 04Request additional information from the customer through a recorded route.
- 05Approve, decline or escalate, recording the reasoning.
- 06The decision, the reasoning and the reviewer are written to the audit trail.
Integrating once and configuring thereafter
- 01Create a sandbox session and exercise the flow with deterministic test identities.
- 02Integrate the API and verify signed webhook handling, including replay attempts.
- 03Configure the verification sequence for the first country and risk level.
- 04Go live for one market.
- 05Add subsequent markets and risk levels through configuration rather than code.
Trust, privacy and security
The model underneath
These are architectural commitments rather than policy statements — each corresponds to something enforced in the system.
- Explicit, recorded consent
- Consent is captured at the point of collection, scoped to a stated purpose, versioned, and retained with the record. Consent for biometric processing is separate from consent for identity verification generally.
- Special-category data handling
- Biometric data is treated as special-category throughout: minimised, encrypted, access-controlled beyond ordinary project membership, retained for a defined period and then deleted rather than archived.
- Explainable decisions
- Every decision can be reconstructed: which checks ran, what each returned, which rules applied, and who overrode what. Risk signals are exposed at the component level so a decision can be defended or appealed.
- No automated rejection without a route back
- A failed check produces a specific, actionable outcome and a path to human review. The product is deliberately not designed to reject a person with no explanation and no appeal.
- Data residency by market
- Where a market requires personal data to remain in-country, the architecture supports it by configuration. Cross-border access is documented with its transfer mechanism.
Architecture and scale
How it is built
- API
- Versioned REST API with idempotency keys, signed webhooks, replay resistance, retry with dead-letter handling, and a sandbox that behaves like production.
- Orchestration
- A durable step engine: verification sequences are data, executed as resumable workflows so a provider outage suspends a case rather than losing it.
- Provider abstraction
- Every document, biometric, registry and data-source provider sits behind a common interface. Adding or replacing one does not touch the domain model or the audit format.
- Evidence store
- Append-only records of checks, decisions and overrides. Documents and biometrics are held in private storage with short-lived signed access and download auditing.
- Review
- A reviewer workspace with queue management, evidence presentation, structured decisions and full attribution.
- Scale and residency
- Verification volume is bursty and campaign-driven. The pipeline is asynchronous throughout, and deployable per region where residency requires it.
Availability
Where you can use it
- Stage
- private beta
- Markets
- United Kingdom, Ghana, Nigeria
- Web
- Not published
- Mobile
- No store listing published
Availability stage, documentation URL and market coverage are awaiting confirmation. Daju Verify does not claim government approval, regulatory accreditation, or guaranteed fraud prevention, and none should be inferred.
Related
Identity and Trust
Identity verification is an orchestration problem, not an integration problem
Teams integrate a verification vendor per country and end up with several incompatible processes and no single audit trail. The fix is to make the sequence itself the thing you build.
Responsible AI
Screen for high-risk processing during discovery, not before launch
A data protection impact assessment discovered two weeks before go-live is a schedule problem. Discovered during discovery, it is a design input.
Industry
Financial services and fintech
Onboarding, money movement and controls that survive an audit.
Industry
Government and public services
Services that must work for everyone, and be defensible afterwards.
Industry
Property and real estate
Long relationships, heavy documents and disputes settled by records.
Next step
Talk to us about Daju Verify
Demonstrations, partnership conversations and pilot enquiries all go to a person, not an automated sequence.