HyperVerge's identity verification flow had to work for a first-time user on a mid-range phone in a Tier 3 city, not just a flagship on 5G. We tested it that way, end to end, on live networks.
A digital KYC app had to hold up for users on older phones and patchy networks in smaller cities. Standard lab testing could not show whether OTP, Aadhaar and VID steps survived those conditions.
Validation & Reliability testing on 60 real devices across three city tiers and live 3G, 4G and 5G networks, with about 30 end-to-end scenarios per device and a 10-language check.
77 issues reported, 39 validated bugs handed to engineering with severity, device, network and repro steps, so each one could be fixed without back-and-forth.
HyperVerge's identity verification app takes a new user from Aadhaar or VID entry through OTP authentication, document capture and e-sign to a completed onboarding. Users arrive on every kind of handset, in every tier of city, and in ten languages.
KYC is already a high-friction moment. If a step stalls or a label reads wrong in someone's own language, they leave, which is why reliability here is a trust question as much as a technical one.
That was the question behind the engagement. The KYC app had to work for everyone, including people in smaller cities on older phones and unreliable connections. Testing in a controlled environment rarely surfaces what those users hit.
The client needed to know whether the OTP arrives, gets validated and moves the user forward, or whether the flow quietly breaks. They also needed to know whether the app holds up across the devices, OS versions and languages that real users bring to onboarding every day.
Before testing began, the team listed every surface to cover: Aadhaar and VID entry, OTP flows, e-sign, T&C acceptance and ASP license validation. 3G conditions and Tier 3 devices were flagged as the highest-risk areas.
Tier 1 cities got 25 devices on 5G, Tier 2 got 23 on 4G and Tier 3 got 12 on 3G, across Android and iOS and brands from Samsung and OnePlus to Vivo, Oppo, Realme and Apple.
Each scenario ran on the actual network assigned to that device, with no simulation. OTP delivery times, Aadhaar API responses and session behaviour were observed under real bandwidth limits.
Aadhaar entry, VID validation, OTP authentication, e-sign and session management ran on all 60 devices, with about 30 scenarios each. That made device-specific failures easy to tell apart from systemic ones.
Each language configuration was reviewed for UI accuracy, T&C content, audio playback and mid-flow switching. Session timeouts, state after interruptions, back-navigation and exit dialogs were tested on top.
All 77 issues were logged with severity, component, device, OS, network condition and reproduction steps. The 39 validated bugs were prioritised so engineering could act on each one directly.
This is Validation & Reliability work: continuously trusted software that performs accurately, safely and reliably in the real world.
Each logged with severity, component, device, OS, network and repro steps.
Prioritised and handed to engineering ready to action.
From OTP flows on 3G to session security and navigation.
UI labels, T&C audio and mid-flow switching all checked.
Missing validation, no autofill, wrong error messages, and OTPs accepted even after the network had failed. All of it appeared only on real 3G.
Some devices refused valid VIDs, cleared input fields on back-navigation and had a refresh button that did nothing.
At key steps the app stopped responding and said nothing, a serious gap for Tier 2 and Tier 3 markets.
A few inaccurate labels, delayed or missing T&C audio, and uneven mid-flow language switching. None were showstoppers, all affect trust.
Sessions stayed active past expected limits and state persisted after the app was closed or interrupted.
Back buttons sometimes skipped the exit confirmation or landed users on the wrong screen during verification.
| Risk before testing | Caught before release |
|---|---|
| OTP flows failing on 3G, blocking onboarding | Every OTP failure identified and documented |
| Valid VIDs rejected, users stuck at step one | Aadhaar and VID issues surfaced with full repro steps |
| App freezing on Airtel 3G with no error shown | 3G instability flagged and prioritised |
| Language and content issues quietly eroding trust | 10-language linguistic check completed, issues logged |
| Sessions staying active after timeout | Session behaviour flagged and queued for remediation |
| 77 issues undetected across the app | 39 validated bugs handed to engineering with full context |
The program gave HyperVerge clear sight of where its KYC journey broke and for whom. Because every device ran the complete flow, the team could separate device-specific failures from systemic ones and fix the right things first. The result is a verification experience that can be strengthened for users in every market, not only the ones with fast phones and fast networks.
OTP failures and app freezes did not appear in standard testing. India's networks vary widely, so a flow that passes in a lab can fail badly on a Tier 3 connection. Run critical journeys on the networks your users have.
Many issues were tied to a specific device or manufacturer, and a narrow device set would have missed them. If your users span city tiers and hardware, your coverage has to span them too.
A slightly off label, missing audio or a broken language switch can tip a hesitant user over the edge. In vernacular markets these affect completion rates, so give them the same attention as functional defects.
For HyperVerge, about 30 end-to-end scenarios on each of 60 real devices turned unknown risk into a prioritised bug list. Tell us which journey worries you most and we will scope it.
In the fast-evolving landscape of app development, ensuring a seamless user experience is paramount. Traditional user testing methods, while effective,...
Read more →
What is AI? Artificial intelligence (AI) is a broad field that includes a variety of techniques and approaches for creating...
Read more →Conducting multiple face recognition trials in different environments and backgrounds to train the AI-based app and validate how it determines...
Read more →Book a 30-minute consultation with an Oprimes AI Trust Specialist. We will map your use case, recommend the right service pillar, and give you a delivery timeline before you commit to anything.
Trusted by 80+ enterprise AI teams across 6 industries. No obligation on first consultation.