Selected highlights from 360+ sprints — a working engine, not a roadmap. Built in Luxembourg.
Core loop
One email in — a complete identity portrait out, in about 90 seconds.
179 OSINT sources
Social, developer and gaming platforms, breach databases, archives and metadata — run in parallel across 11+ categories.
11-axis behavioral fingerprint
A signature that persists across accounts, even when infrastructure rotates.
Identity graph
Accounts, breaches, domains and locations linked into personas, with confidence propagated across the graph.
Relationship graph
See how the identities and organizations in a workspace connect — a people-and-entity view layered over the raw findings.
Dual scoring
Exposure (footprint size) and threat (breach severity) measured as independent signals.
Breach & leak intelligence
Coverage across multiple breach and leak databases, with timelines.
Sanctions & PEP screening
OFAC, EU and UN watchlists, plus corporate officer and directorship records.
Legal records
Public court and registry records across the US, France and the UK.
Public exposure
Global news and media monitoring tied to the resolved identity.
Phone & wallet extraction
Secondary identifiers surfaced from breach data and cross-referenced.
Persona clustering
Multiple identities resolved, grouped and compared automatically.
Correct-person binding
A finding only attaches to a person when the match corroborates across signals — so a page that merely name-drops a company, or shares a common first name, never resolves to the wrong individual.
Per-finding provenance
Source, confidence tier, cross-verification and first/last-seen on every result.
SOC exposure API (SIEM/SOAR)
Machine-readable identity-exposure lookups for security pipelines — sourced evidence and corroboration, key-based access, never a singular verdict.
On-prem, zero-LLM enrichment
The exposure path is fully deterministic — no language model in the loop; any model-assisted extraction stays inside your environment.
AI-assisted extraction · on-prem
On the discovery path, a local language model reads prose-heavy pages that pattern-based parsers miss — surfacing candidate people as review-only leads, never an automatic resolution. It runs inside your environment; no identity data is sent to a third-party LLM API.
Generative identity avatars
Deterministic pixel-art portraits from the identity graph — zero GPU, zero external API.
PDF identity reports
A 5-page, plan-tiered report for every scan.
GDPR-native by design
Self-scan is always free; third-party scans require documented consent.