Talk to the Team

Tell us about your stack and the privacy problems you're trying to solve. We typically respond within one business day.

Prefer email? support@philterd.ai

Prefer to skip the form? Pick a time on our calendar →
or send a message

Please do not enter PII or PHI in this form. If you need to share an example, use a sanitized one.

Compliance

Compliance Matrix

Which Philterd products address which regulations: HIPAA, GDPR, PIPEDA, Law 25, CCPA, PCI DSS, GLBA, FERPA, SOX, FedRAMP, and more. For your security review.

This matrix is provided for informational purposes only and does not constitute legal or compliance advice. Consult with your security team and legal counsel to determine whether Philterd's products meet the specific requirements of your regulatory environment.

Regulation coverage

Every Philterd product is self-hosted: your data never leaves your infrastructure. The matrix below maps each regulation to the products and capabilities that address it, including PHI redaction for HIPAA.

RegulationScopeProductsKey capabilitiesPolicy
HIPAA Safe Harbor
45 CFR 164.514(b)(2)
All 18 PHI identifiers must be removed for data to qualify as de-identified.Philter Phileas PhEye ArbiterDetection of PHI identifiers including names, dates, geographic subdivisions, ages over 89, MRNs, and other unique identifiers. Healthcare NLP lens for clinical text. Human review via Arbiter.Yes
HIPAA Expert Determination
45 CFR 164.514(b)(1)
Qualified statistician certifies re-identification risk is very small.Philter Phileas Philter DiffuseRedaction of direct identifiers. Differential privacy for aggregate analytics on residual data.Configurable
GDPR
EU 2016/679
Personal data of EU data subjects must be processed lawfully; right to erasure.Philter Phileas Phinder Philter AI ProxyPII discovery across data stores (Phinder). Redaction of personal data categories: names, addresses, national IDs, dates of birth, email, phone. Prompt/response redaction for LLM workloads (AI Proxy).Configurable
PIPEDA
S.C. 2000, c. 5
Personal information of individuals in Canada collected, used, or disclosed in commercial activities by private-sector organizations.Philter Phileas Phinder Philter AI ProxyRedaction of personal information categories: names, SINs, addresses, dates of birth, health information, financial details, contact information. Phinder scans data stores for personal information. Self-hosted deployment keeps data within Canada.Configurable
Law 25 (Quebec)
An Act respecting the protection of personal information in the private sector
Personal information of Quebec residents; strict data residency requirements and bilingual processing obligations.Philter PhileasSelf-hosted deployment satisfies data residency requirements (deploy in a Canadian AWS, Azure, or GCP region). Multilingual redaction handles French and English text with a policy per language. Canadian identifier detection (SIN, provincial health card patterns).Configurable
CCPA / CPRA
Cal. Civ. Code 1798.100
California consumers' personal information; right to delete.Philter Phileas PhinderSame detection and redaction capabilities as GDPR. Phinder scans for personal information across S3, GCS, Azure Blob, and local filesystems.Configurable
PCI DSS
v4.0, Req 3.2-3.4
Cardholder data: PAN, CVV, track data. PAN must be rendered unreadable when stored.Philter Phileas PhinderCredit card number detection with Luhn validation. Masking PAN to last 4 digits (Req 3.4). Full redaction of CVV/CVC (Req 3.2). PII discovery in logs, transcripts, and storage.Yes
GLBA
Gramm-Leach-Bliley Act
Nonpublic personal information (NPPI) of financial institution customers.Philter Phileas PhinderDetection of NPPI: SSNs, account numbers, income data, credit history references. Discovery scanning across document stores.Configurable
FRBP 9037
Bankruptcy Rule
Court filings must show only last 4 of SSN/taxpayer ID, year of birth, initials of minors, last 4 of financial accounts.Philter Phileas ArbiterLast-4 masking for SSNs and account numbers. Year-only date truncation. Initials for minor names. Attorney QC review via Arbiter.Yes
FRCP 5.2
Federal Civil Procedure
Same redaction requirements as FRBP 9037, applied to all federal civil and criminal filings.Philter Phileas ArbiterSame capabilities as FRBP 9037. Applies to electronically filed documents in all federal courts.Yes
FERPA
20 U.S.C. 1232g
Student education records: names, IDs, grades, disciplinary records.Philter PhileasDetection of student names, student IDs, dates of birth, addresses, and education-record-specific identifiers.Configurable
SOX
Sarbanes-Oxley Act, Section 802
Financial records and audit trails must be preserved but sensitive data must be controlled.Philter Phileas PhieldRedaction of PII in financial documents. Phield monitors PII trends across financial data pipelines to detect anomalous flows.Configurable
FedRAMP
NIST 800-53
Cloud services used by federal agencies must meet NIST 800-53 controls.PhilterDeployable in FedRAMP-authorized environments: AWS GovCloud, Azure Government, Google Distributed Cloud Hosted. Self-hosted with no outbound dependencies.N/A
CMMC
Cybersecurity Maturity Model
Defense contractors must protect Controlled Unclassified Information (CUI).PhilterAir-gapped deployment. No external API calls. Self-contained Docker images.N/A
ITAR
International Traffic in Arms
Technical data related to defense articles must be controlled.PhilterAir-gapped deployment in classified environments. No data leaves the perimeter.N/A

Product capabilities by compliance function

FunctionProductWhat it does
Detect PII in textPhilter Phileas30+ entity types. Pattern-based detection (regex, checksums, format validators) plus NLP-based detection via PhEye lenses.
Redact / mask / encryptPhilter PhileasMultiple strategies per entity type: full redaction, masking to last N characters, year-only date truncation, initials, format-preserving encryption, synthetic value replacement.
Discover PII across data storesPhinderCrawls S3, GCS, Azure Blob, and local filesystems. Reports entity types and counts per file or object.
Monitor PII trendsPhieldReceives PII type counts via API or Kafka. Stores in MongoDB time-series or in-memory. Alerts via PagerDuty or Slack when counts deviate from established trends.
Human review of automated redactionsArbiterAccept, override, or exempt each detection. Structured exemption codes. Per-reviewer throughput and audit reporting.
Benchmark redaction accuracyPhilter ScopePrecision, recall, F1 per entity type against gold-standard test data. Entity type confusion matrix. CI/CD integration for regression detection.
Differentially private analyticsPhilter DiffuseFormal epsilon-budget differential privacy for aggregate queries (counts, sums, averages) on PII telemetry. Membership-inference resistant.
Redact PII from LLM promptsPhilter AI ProxyDrop-in proxy for OpenAI, Anthropic, Gemini, and Ollama. Redacts PII from prompts before the model sees them.
Build and edit redaction policiesRedaction Policy EditorVisual, no-code policy builder. Exports valid Philter-compatible policies.

Deployment models

All Philterd products are self-hosted. Your data stays in your infrastructure.

DeploymentProducts supportedNotes
Cloud VPC (AWS, GCP, Azure)AllStandard deployment. Cloud marketplace listings available for Philter.
AWS GovCloud / Azure GovernmentPhilterFedRAMP Moderate and High.
Air-gapped / disconnectedPhilterSelf-contained Docker images. No outbound network required.
On-premisesAllDocker or bare-metal. No SaaS dependency.

Pre-built redaction policies

Ready-to-use policy files for common regulations. Available in the policy library and on GitHub.

PolicyRegulationWhat it covers
HIPAA Safe HarborHIPAA 45 CFR 164.514(b)(2)All 18 Safe Harbor identifiers with appropriate redaction strategies per identifier type.
PCI DSS Scope ReductionPCI DSS Req 3.2-3.4PAN masking to last 4, full CVV/CVC redaction.
Rule 9037 BankruptcyFRBP 9037SSN/taxpayer ID to last 4, dates to year only, minor names to initials, financial accounts to last 4.
General PurposeMultipleBroad entity detection for names, SSNs, credit cards, dates, addresses, phone numbers, emails.
LLM Training DataAI/ML governanceComprehensive PII stripping for training corpus preparation.
Clinical Notes for ResearchHIPAAOptimized for narrative clinical text with healthcare NLP lens.

Community-contributed policies are welcome via pull request.