Frequently asked questions.
How Sign3 fits your stack, how fast it goes live, how your data is protected, and what results look like in production.
Sign3 is an AI-native customer intelligence platform for Indian financial institutions. It captures first-party signals from the device, digital footprint, behaviour, location, image, and SMS, and resolves them into a unified customer persona that makes onboarding, fraud, credit, and compliance decisions sharper.
No. Sign3 sits on top of existing infrastructure. It adds the intelligence layer that makes your current systems sharper. No replacement, no migration. Bureau, KYC, rule engine, and TMS continue to run as they are. Sign3 enriches them.
Yes. Sign3's pre-built scores are decision-ready and require no internal modelling. The platform returns risk scores, flags, and recommended actions (allow, warn, block) that plug directly into any rule engine or BRE. Institutions with data science teams can consume raw signals into their own models. Institutions without them use the scores as they are.
Sign3 is ISO 27001:2022 certified. All data is encrypted at rest (AES-256) and in transit (TLS 1.2+). On-premise deployment is available with no external connectivity post-deployment. Sign3 operates as a data processor, not a data controller. Data residency is India by default.
Yes. Digital Footprint relies on publicly available data exempt under DPDP Section 3(c)(ii). Device Intelligence collects non-PII data. Sign3 is GDPR compliant. VAPT reports and security documentation are available upon request.
API-only modules (Digital Footprint, Location Intelligence, Image Intelligence) go live in 1-2 weeks with no app-side changes. SDK modules (Device Intelligence, Behavioural Biometrics) integrate in 3-4 weeks. Default models are production-ready from day one. Custom portfolio calibration takes 2-3 weeks with historical data.
In production: 73% of mule accounts flagged at onboarding at a banking client. 65% of NPAs concentrated in the bottom 5% of scored applicants at a credit card portfolio. 503 fraud rings identified from a single analysis window at a lending client. These are outcomes on data that had already cleared existing KYC and rule-based checks.
Yes. Sign3's alternate data signals provide a credit view where bureau returns nothing. One lending partner saw disbursals grow from ₹35 Cr to ₹80 Cr per month after integrating Sign3 signals into underwriting, with no corresponding rise in defaults.
Yes. Full on-premise deployment is available with no external connectivity post-deployment. Sign3 has established integration patterns with Finacle, BaNCS, and Flexcube. This is the preferred option for several banking clients.
Sign3 offers pilot and POC deployments. Share historical data (phone numbers, emails, outcome labels) and Sign3 will run its models against your portfolio, share the results, and let the numbers speak. Minimum data for calibration: 20K-50K historical records.