What happens when a returning user opens your app, starts a verification, and then closes it halfway through? That single moment, the resume, quietly decides whether you keep a customer or lose one to a spinning liveness spinner. Most teams obsess over first-time onboarding and treat re-verification as an afterthought. Then renewal season arrives, and the cracks show.
This piece looks at how Onfido and Sumsub handle resume and re-verification flows, where the practical differences bite, and what to plan for as 2027 approaches. One honest note before we start: the phrase “Igt Smoke” in the original query isn’t a term with a clear public definition. It may be a typo, an internal codename, or niche jargon. So this article focuses on the part that can actually be researched and compared: the resume workflow itself. (Keyboard foam mod sound differences actually changes)
Why Resume Flows Matter More Than You Think
A resume flow is the ability to pick up a partially completed identity check across sessions, devices, or time gaps. Sounds simple. It rarely is. The moment a user drops off mid-verification, several invisible clocks start ticking: session tokens expire, liveness captures go stale, document images drift out of date, and regulatory freshness windows creep closer.
The regulatory backdrop is the reason this matters. Periodic review and ongoing monitoring are baked into international anti-money-laundering standards, and the FATF Recommendations set out the customer due diligence expectations that most national regimes copy. Re-verification isn’t a product nicety. It’s a compliance obligation with a user-experience wrapper around it.
Where teams go wrong: they treat resume as a UX feature owned by the frontend team. In reality, it’s a state machine that touches identity proofing standards, session security, and audit logging. The NIST Digital Identity Guidelines on identity proofing explain how proofing and re-proofing are structured, and reading them makes it obvious why “just save the session” is not a design. (Identity verification service)
Vendors in Plain Terms
Onfido is a UK-founded identity verification provider, known for document and biometric checks in KYC and AML onboarding. It was acquired by Entrust, announced in 2024, which matters for anyone planning a 2027 roadmap: product naming, documentation, and roadmap continuity may shift as the two organizations integrate.
Sumsub is an identity verification and compliance platform with international operations. Its positioning leans toward a broad compliance suite, with re-verification and ongoing monitoring marketed as part of the platform rather than a bolt-on.
Both vendors offer resume and re-verification workflows. Both change their feature names and mechanics frequently. Anyone claiming a definitive, permanent feature-by-feature comparison is probably reading a stale sales deck. Treat vendor documentation as the source of truth and re-check it before you sign.
Where the Workflow Differences Actually Show Up
Strip away the marketing and resume workflows differ along a handful of practical axes. Here’s the comparison that matters operationally.
| Workflow dimension | What to check on Onfido | What to check on Sumsub |
|---|---|---|
| Session persistence | How long a partial check stays resumable, and whether it survives a device switch | Whether resume tokens are tied to a single applicant profile or a session |
| Liveness re-checks | Whether a returning user must redo liveness or can reuse a prior capture | Freshness window for biometric captures before a re-check is forced |
| Document re-capture | Rules for expired or near-expiry documents | Whether previously submitted documents are retained and re-scored |
| Re-verification triggers | Manual initiation versus rule-based triggers | Rule engine for periodic review and risk-based re-checks |
| Audit trail | What the dashboard logs for a resumed session | Granularity of the audit log for compliance evidence |
| Failure recovery | Behavior when a resume attempt fails mid-flow | Fallback path when a returning user’s stored data is rejected |
None of these rows has a universal winner. The right answer depends on your user base, your regulator, and how often your users return.
Session persistence and the device-switch problem
This is the most common source of support tickets. A user starts verification on a phone, gets interrupted, and finishes on a laptop. If the resume token is bound to the original session, the flow restarts from zero. Both vendors support cross-session resume in some form, but the mechanics differ, and the details are exactly the kind of thing that changes between releases. Test it yourself with a real device switch before you commit. Don’t take a sales engineer’s word for it.
Liveness re-checks and freshness windows
Biometric captures age. A liveness check from eight months ago is not the same evidence as one from last week, and regulators know it. The practical question is whether the vendor lets you configure a freshness window or hard-codes one. If it’s hard-coded and your regulator wants annual re-verification, you may end up forcing full re-checks more often than necessary, which is friction you’re paying for in churn.
Document re-capture on expiry
Passports expire. Residence permits get renewed. A good resume flow notices a document is near expiry and prompts a re-capture at the right moment rather than failing the user at checkout. Whether that logic lives in the vendor’s platform or in your own orchestration layer is a design decision worth making explicitly. Many teams assume the vendor handles it and discover during an audit that nobody did.
The Compliance Layer Nobody Wants to Read
Periodic review requirements are the reason resume flows exist in the first place. Under the FATF Recommendations, customer due diligence includes ongoing monitoring, and most national supervisors translate that into periodic re-verification for higher-risk customers. The EU’s electronic identification framework, the eIDAS Regulation, adds another layer for cross-border identity assurance, which matters if you operate across European markets.
The practical implication: your resume workflow is evidence. When a supervisor asks why a customer wasn’t re-verified on schedule, “the session expired” is not an answer. You need a logged trigger, a logged attempt, and a logged outcome. Both vendors provide audit trails, but the granularity differs, and you should confirm what gets exported versus what stays locked in the dashboard. (Use claude writing practical workflow that actually works)
What Actually Works in Practice
After enough implementations, patterns emerge. Here’s what tends to hold up.
- Map your re-verification triggers before you evaluate vendors. Write down every reason a user needs to re-verify: document expiry, risk score change, regulatory cycle, account dormancy. Then ask each vendor how they support each trigger.
- Test resume across devices and time gaps. Start a check on one device, wait 24 hours, finish on another. Do it again with a 30-day gap. Document what happens.
- Confirm what the audit log actually contains. Export it. Read it. If it doesn’t show the resume attempt and its outcome, it won’t satisfy a supervisor.
- Decide where orchestration lives. Vendor platform or your own layer. Both work. Mixing them without a clear owner is how gaps appear.
- Build a fallback path. When resume fails, the user should land somewhere sensible, not a dead end. A support handoff beats a blank screen.
The most common mistake is treating resume as a checkbox during procurement. The second most common is assuming the vendor’s default freshness windows match your regulator’s expectations. They usually don’t, and the gap only surfaces during an audit or a customer complaint.
Planning for 2027: What to Watch
Nobody can predict a 2027 roadmap with confidence, and any article claiming otherwise is guessing. What you can do is watch the signals.
The Entrust acquisition of Onfido is the biggest one. Integration periods typically bring documentation changes, product renaming, and occasionally deprecation of overlapping features. If you’re on Onfido today, ask your account team directly what happens to resume and re-verification features through the integration. Get it in writing.
On the Sumsub side, watch how the compliance suite expands. Broader platforms tend to add configurability, which is good for regulated buyers and bad for teams that want simple defaults. (What is biometrics?)
Regulatory signals matter too. Periodic review expectations tighten over time, and cross-border identity frameworks keep evolving. A resume workflow that’s compliant today may need a shorter freshness window in two years. Build for configurability now and you won’t be rebuilding later.
If you’re also weighing broader tooling decisions alongside identity infrastructure, it helps to see how other workflow choices play out in practice. Comparisons like the Obsidian vs Logseq local-first notes breakdown by workflow show the same principle: the right pick depends on how people actually move through the tool, not on a feature list.
Common Mistakes and How to Avoid Them
Assuming resume is automatic. It isn’t. It’s a configured behavior with limits. Ask for the limits.
Ignoring the device-switch case. Mobile-first users switch devices constantly. Test it.
Skipping the audit export test. A dashboard view is not evidence. An export is.
Letting the vendor own your re-verification schedule. Your regulator’s calendar is yours, not theirs. Own the triggers.
Forgetting the failure path. A failed resume should route somewhere helpful. Design it.
Conclusion
The Onfido versus Sumsub resume question doesn’t have a single winner, and anyone selling you one is oversimplifying. What matters is the fit between your re-verification triggers, your regulator’s expectations, and how each vendor’s session persistence, freshness windows, and audit trails actually behave. Test the device-switch case, export the audit log, and get the Entrust integration roadmap in writing before 2027 planning gets serious. The workflow differences are real, but they only matter once you’ve mapped your own requirements first.
Frequently Asked Questions
What does “resume” mean in identity verification?
A resume flow lets a user pick up a partially completed identity check later, on the same or a different device, instead of starting over. It depends on session persistence, token expiry rules, and how long biometric and document captures stay valid.
Is “Igt Smoke” a known term in identity verification?
No clear public definition exists. It may be a typo, an internal codename, or niche jargon. Anyone building a comparison around it should confirm what it refers to before relying on it.
Did Entrust acquire Onfido?
Yes, the acquisition was announced in 2024. For planning purposes, the key question is how product naming, documentation, and resume features evolve through the integration. Ask the vendor directly and get answers in writing.
Which vendor handles re-verification better?
Neither wins universally. Onfido and Sumsub both support re-verification, but the mechanics differ in session persistence, liveness freshness windows, document re-capture logic, and audit granularity. The better fit depends on your triggers and your regulator.
How often should users be re-verified?
It depends on your jurisdiction and risk profile. International anti-money-laundering standards require ongoing monitoring and periodic review, but the specific cadence comes from your national supervisor and your own risk assessment.
What should I test before choosing a vendor?
Test resume across devices with a 24-hour gap and a 30-day gap. Export the audit log and confirm it records resume attempts and outcomes. Confirm how document expiry triggers re-capture. And ask what happens to the feature set through any corporate integration.
Document the test results. A short internal memo beats a vendor’s marketing deck when your auditor asks why you chose one platform over the other. Record the date, the scenario, and the observed behavior for each row in the table above. That record becomes your defense if a supervisor later questions your re-verification cadence, and it costs an afternoon to produce. The teams that get burned are the ones who evaluated on feature checklists and never wrote down what they actually saw when a real user resumed a real check on a real device.
