Identity Intelligence

People Score: Identity Confidence at a Glance

A composite AI-powered score that aggregates phone reputation, email signals, social footprint, and name validation into a single number — telling you instantly how real and trustworthy an identity is.

AI + machine learning

4 signal types combined

Single API call

Get Started Free API Docs

High confidence

Real identity

4 signals merged

People Score

IRBIS AI Validation Engine

Calculating…

J. Anderson

+1 (415) 555-2671 · janderson@gmail.com

Verified

Phone signal

Email depth

Social footprint

Name validation

Signal breakdown

Four signals. One composite score.

The People Score is not a single lookup — it's a weighted combination of four independent identity signals processed by the IRBIS AI engine.

Phone · 25%

Phone Reputation

Carrier validation, line type (mobile/VoIP), activity status, and spam reputation contribute to the phone signal layer.

Email · 30%

Email Identity Depth

Deliverability, social footprint age, platform count, and breach history measure how established and authentic the email identity is.

Social · 25%

Social Footprint

Presence across social networks, consistency of identity across platforms, and linked username patterns build the social signal.

Name · 20%

Name Validation

AI cross-references the provided name against phone data, messenger registrations, and caller ID records to validate identity consistency.

Score interpretation

What each score range means

The score maps directly to an identity confidence level — giving your system a clear, actionable signal to route, approve, or escalate.

80–100

High Confidence

Strong multi-signal match. Real person with established digital identity. Safe to proceed with minimal friction.

60–79

Moderate Confidence

Identity signals are largely consistent but incomplete. Consider light additional verification for high-value actions.

40–59

Low Confidence

Thin or inconsistent identity signals. Recommend secondary verification before proceeding with onboarding or transactions.

0–39

High Risk

Very limited or contradictory identity signals. Likely a synthetic, fake, or bot-generated identity. Flag for review or block.

Who uses it

Built for any team that needs to know how real a person is

One score. Actionable across fraud prevention, KYC, onboarding, marketing, and risk orchestration workflows.

KYC & Compliance

Replace lengthy manual checks with a single score that reflects the full identity picture. Flag thin identities at onboarding before they become a compliance risk.

Finance & Banking

Score new account applicants in real time. Integrate the People Score into your AML and fraud risk engine to weight synthetic identity risk automatically.

E-Commerce

Score buyers and sellers at registration. Low-confidence accounts can be required to verify further before completing high-value transactions or marketplace listings.

Lead Scoring

Enrich CRM leads with a People Score at the point of capture. Prioritize outreach to high-confidence identities and filter out bot-generated or fake form submissions.

SaaS & Digital Platforms

Block multi-account abuse and fake registrations at signup. Use the score threshold as a gate for trial access, promotional credits, or high-trust features.

Risk Orchestration

Feed the People Score directly into your risk engine as a weighted input signal — alongside device fingerprint, transaction history, and behavioral signals.

API response fields

Every field returned by the People Score API

The composite score plus individual signal breakdowns — all in one normalized JSON response.

Composite People Score (0–100)Always returned

Phone signal score (0–100)Always returned

Email depth score (0–100)Always returned

Social footprint score (0–100)Always returned

Name validation score (0–100)Always returned

Confidence level & decision labelAlways returned

people-score-response.json

// POST /api/developer/people_score { "id": 5519203, "status": "completed", "people_score": 87, "confidence": "high", "decision": "real_person", "signals": { "phone": 88, "email": 94, "social": 76, "name": 91 }, "name": "J. Anderson", "phone": "+14155552671", "email": "janderson@gmail.com" }

FAQ

Common questions

What is the People Score?

The People Score is a composite identity confidence score (0–100) that aggregates four signal types — phone reputation, email depth, social footprint, and name validation — into a single number processed by the IRBIS AI engine.

What inputs does the API require?

At minimum, a phone number or email address. Adding a full name significantly improves score accuracy. All four inputs (phone, email, name, and optional additional identifiers) can be submitted together for the most precise composite score.

How is the score calculated?

IRBIS processes each signal independently — phone (25%), email (30%), social (25%), and name (20%) — then the AI engine weights and merges them into a composite score. Individual signal scores are also returned so you can inspect each component.

What score threshold should I use?

Most customers use 70+ as a green light, 50–69 as a soft review trigger, and below 50 as a flag for additional verification or block. The right threshold depends on your risk appetite and use case — our team can help calibrate this.

Is the People Score GDPR-compliant?

Yes. ESPY is GDPR-compliant and ISO 27001 certified. The People Score uses publicly available data sources and carrier-grade registries. No personally identifiable data is stored beyond the scope of the API call.

How fast does the score return?

The People Score uses an async job model. Submit the request, receive a job ID, and poll for results — typically completing within 20–45 seconds as all four signal pipelines run in parallel.