AnonShield On-premise PII pseudonymization, built for CSIRTs.

Redact sensitive entities before sharing security data. Zero cloud, zero persistence: everything runs inside your network, processed in-memory and deleted on the spot.

Live demo runs entirely in your browser

Mode

Reversible: each entity becomes a stable token. With the secret key you can re-identify and keep correlations across documents.

Input editable: type to test

Output 10 entities redacted

Meeting notes: Q4 Security Review Attendees: Sarah Chen ([EMAIL-2g07bt]), Marcus Rodriguez Phone: [PHONE-b0hxs5] · Alt: [PHONE-6i11m8] Action items: • Card on file: [CREDIT·CARD-fj28wp] (exp. 09/27) • Migrate server [IP·ADDRESS-ev73g9] → new host at [IP·ADDRESS-63vl0e] • Patch [CVE·ID-aa8y9z] before Friday • API token: [AUTH·TOKEN-x5p27n] • Leaked URL: [URL-e1ebcm] • File hash: [HASH-s5gyw4]

PHONE 2

IP_ADDRESS 2

EMAIL 1

CREDIT_CARD 1

CVE_ID 1

AUTH_TOKEN 1

URL 1

HASH 1

Mode trade-offs

Reversible Yes, with key

Correlation Preserved

Privacy High

Launch the app View the slides No account · No upload · Open source

How it works

Five stages, zero data left behind Click any stage to inspect it

Ingest Read any format

Detect NER + regex

Hash HMAC-SHA256

Replace Typed tokens

Deliver Download & purge

Nothing is written to disk. The original is processed in memory and discarded the moment your file is ready.

738× faster, over 92 h reduced to under 10 min

94.2% F1 score

96.4% recall

550MB processed in under 10 minutes

0 bytes sent to the cloud

1MB max file size in this demo

Supported formats

TXTCSVJSONJSONLPDFDOCXXLSXXMLZIPPNGJPGTIFFBMPWEBP

Recognition

Best Artifact Award at SBRC 2026

AnonShield was selected as the best artifact of the entire conference, and earned all four SBC reproducibility seals.

SBC reproducibility seals

View official results ↗ Read the paper ↗

Research lineage

Three generations, one mission

From AnonLFI to AnonShield: a peer-reviewed evolution toward scalable, on-premise pseudonymization for CSIRTs.

SBSeg 2025 AnonLFI v1.0

Anonimização de Incidentes de Segurança com Reidentificação Controlada

Carolina Tompsen Bandel (SENAC EAD), João Pedro Ramires Esteves (UFU), Kalian Pereira Guerra (UFRGS), Leandro M. Bertholdo (UFRGS), Diego Kreutz (UNIPAMPA), Rodrigo S. Miani (UFU)

Natural-language incident reports, anonymized for safe LLM use with controlled re-identification

100% Precision 97.38% efficacy 763 incidents Controlled re-identification NER + regex On-premise

Paper ↗ Code ↗

WRSeg / ERRC 2025 AnonLFI v2.0

AnonLFI 2.0: Extensible Architecture for PII Pseudonymization in CSIRTs with OCR and Technical Recognizers

Cristhian Kapelinski (UNIPAMPA), Douglas Lautert (UNIPAMPA), Beatriz Machado (UNIPAMPA), Diego Kreutz (UNIPAMPA)

Reversible pseudonymization with OCR and technical recognizers, preserving XML and JSON structure

100% Precision 92.13% F1 (OpenVAS XML) 76.5% F1 (OCR / PDF) Reversible (HMAC-SHA256) XML / JSON structure preserved On-premise

Paper ↗ Code ↗

SBRC 2026 AnonShield ← You are here

AnonShield: Scalable On-Premise Pseudonymization for CSIRT Vulnerability Data

Cristhian Kapelinski (UNIPAMPA), Douglas Lautert (UNIPAMPA), Beatriz Machado (UNIPAMPA), Isadora Garcia Ferrão (UBO), Diego Kreutz (UNIPAMPA)

GPU-accelerated, streaming pseudonymization of network vulnerability scans at scale

94.2% F1 96.4% Recall 738× faster 92h to <10 min 550 MB / 70,951 records GPU-accelerated NER On-premise

Paper ↗ Code ↗

The team

Built by researchers, for CSIRTs

Cristhian Kapelinski UNIPAMPA

Douglas Lautert UNIPAMPA

Beatriz Machado UNIPAMPA

Diego Kreutz UNIPAMPA

Isadora G. Ferrão UBO

Carolina Tompsen Bandel SENAC EAD

João Pedro Ramires Esteves UFU

Kalian Pereira Guerra UFRGS

Leandro M. Bertholdo UFRGS

Rodrigo S. Miani UFU

Developed at UNIPAMPA in collaboration with UBO (Université de Bretagne Occidentale), UFU, UFRGS, and SENAC.

Ready to anonymize your data?

Run it on-premise in seconds. No account, no upload, no data ever leaves your network.

Launch the app [email protected]