queryable-encryption-workshop

Trigger

Use this skill when demonstrating Queryable Encryption, mapping compliance requirements, or building industry-specific security demos.

Trigger phrases:

• "Build QE demo for [healthcare/finance/government]"
• "Map QE to HIPAA/PCI-DSS/GDPR"
• "Generate encryption demo"
• "Show QE compliance patterns"

Overview

Queryable Encryption (QE) is MongoDB's client-side field-level encryption that allows encrypted search. Different industries have different compliance needs, data sensitivity levels, and use cases.

This skill generates industry-customized demos that map QE features to specific compliance requirements.

Not for: General encryption education. This is about vertical-specific demos with compliance context.

How to Use

Quick Start

1. Analyze vertical requirements:

   python scripts/vertical_analyzer.py --vertical healthcare --output analysis.json

2. Generate demo code:

   python scripts/demo_generator.py analysis.json --output demo/

3. Map compliance requirements:

   python scripts/compliance_mapper.py --vertical healthcare --framework HIPAA --output compliance.md

Python Tools

• scripts/vertical_analyzer.py — Analyze industry data sensitivity patterns
• scripts/demo_generator.py — Generate vertical-specific QE demo code
• scripts/compliance_mapper.py — Map QE to compliance frameworks

Reference Docs

• references/qe-compliance-guide.md — QE mapped to HIPAA, PCI-DSS, GDPR
• references/industry-patterns.md — Common patterns by vertical

Templates & Assets

• assets/healthcare-schema.json — Patient data with PHI fields
• assets/finance-schema.json — Financial data with PII/PCI fields
• assets/government-schema.json — Citizen data with SSN/TIN fields

Industry Verticals

Healthcare (HIPAA)

Sensitive data:

• PHI (Protected Health Information)
• SSN, date of birth, medical record numbers
• Diagnoses, prescriptions, lab results

QE approach:

• Equality queries: SSN, medical record number (exact lookups)
• Range queries: Date of birth (age-based queries)
• Unencrypted: Non-PHI (names for display, city for analytics)

Example schema:

{
  patientId: "P12345",  // Deterministic (equality search)
  ssn: "***-**-6789",   // Deterministic (encrypted, searchable)
  dob: "1980-05-15",    // Range (encrypted, age queries)
  diagnosis: "...",     // Random (encrypted, not searchable)
  fullName: "John Doe", // Plaintext (not PHI for display)
  city: "Boston"        // Plaintext (analytics)
}

Compliance mapping:

• HIPAA §164.312(a)(2)(iv): Encryption at rest ✅
• HIPAA §164.312(e)(2)(ii): Encryption in transit ✅
• HIPAA §164.308(a)(4): Access controls ✅ (key management)

Finance (PCI-DSS)

Sensitive data:

• Credit card numbers (PAN)
• CVV, expiration dates
• Account numbers, routing numbers
• SSN, transaction details

QE approach:

• Deterministic: Last 4 digits of card (customer lookup)
• Random: Full card number (store but never query)
• Unencrypted: Transaction amounts (reporting), merchant names

Example schema:

{
  customerId: "C12345",      // Deterministic
  cardLast4: "4242",         // Deterministic (searchable)
  cardNumberFull: "****",    // Random (encrypted, not searchable)
  expirationDate: "12/25",   // Random
  transactionAmount: 99.99,  // Plaintext (analytics)
  merchantName: "Store Inc"  // Plaintext
}

Compliance mapping:

• PCI-DSS Req 3.4: Encryption of PAN ✅
• PCI-DSS Req 3.5.1: Key management ✅
• PCI-DSS Req 3.6.4: Cryptographic key changes ✅

Government (FedRAMP, FISMA)

Sensitive data:

• SSN, TIN (Tax ID)
• Clearance levels
• Citizen records
• Classified information

QE approach:

• Deterministic: SSN (citizen lookup)
• Range: Security clearance level (range queries)
• Random: Classified notes
• Unencrypted: Public records, statistics

Example schema:

{
  citizenId: "CIT12345",    // Deterministic
  ssn: "***-**-6789",       // Deterministic (searchable)
  clearanceLevel: 3,        // Range (encrypted, level queries)
  classifiedNotes: "...",   // Random (encrypted, not searchable)
  state: "MA",              // Plaintext (analytics)
  registrationDate: "2020"  // Plaintext
}

Compliance mapping:

• FISMA: Encryption requirements ✅
• FedRAMP: Key management ✅
• NIST 800-53: Access controls ✅

QE Encryption Types

1. Deterministic (Equality Queries)

Use for: Exact match searches (SSN, email, account number)

Properties:

• Same plaintext → same ciphertext (searchable)
• Supports: $eq, $in, $ne
• Not secure for low-cardinality fields (gender, state)

Example:

{
  ssn: {
    encrypt: {
      keyId: "key1",
      algorithm: "AEAD_AES_256_CBC_HMAC_SHA_512-Deterministic",
      bsonType: "string"
    }
  }
}

2. Range (Range Queries)

Use for: Numeric ranges, date ranges (age, salary, dates)

Properties:

• Supports: $gt, $gte, $lt, $lte, $range
• More secure than deterministic
• Requires min/max bounds

Example:

{
  dateOfBirth: {
    encrypt: {
      keyId: "key2",
      algorithm: "Range",
      bsonType: "date",
      min: ISODate("1900-01-01"),
      max: ISODate("2024-12-31"),
      sparsity: 2
    }
  }
}

3. Random (No Queries)

Use for: Sensitive data never queried (full card numbers, notes)

Properties:

• Same plaintext → different ciphertext (most secure)
• Not searchable
• Best for high-sensitivity fields

Example:

{
  creditCardNumber: {
    encrypt: {
      keyId: "key3",
      algorithm: "AEAD_AES_256_CBC_HMAC_SHA_512-Random",
      bsonType: "string"
    }
  }
}

Key Management Patterns

Pattern 1: AWS KMS (Production)

Best for: Production deployments, compliance audits

Setup:

const clientEncryption = new ClientEncryption(client, {
  keyVaultNamespace: "encryption.__keyVault",
  kmsProviders: {
    aws: {
      accessKeyId: process.env.AWS_ACCESS_KEY_ID,
      secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY
    }
  }
});

Compliance: FIPS 140-2 validated, audit trails, key rotation

Pattern 2: Local Key (Demos)

Best for: Workshops, POCs, local development

Setup:

const localKey = crypto.randomBytes(96);
const clientEncryption = new ClientEncryption(client, {
  keyVaultNamespace: "encryption.__keyVault",
  kmsProviders: {
    local: {
      key: localKey
    }
  }
});

Warning: Not for production (key stored in code)

Pattern 3: Azure Key Vault

Best for: Azure deployments, Microsoft shops

Setup:

const clientEncryption = new ClientEncryption(client, {
  keyVaultNamespace: "encryption.__keyVault",
  kmsProviders: {
    azure: {
      tenantId: process.env.AZURE_TENANT_ID,
      clientId: process.env.AZURE_CLIENT_ID,
      clientSecret: process.env.AZURE_CLIENT_SECRET
    }
  }
});

Demo Structure

Phase 1: Unencrypted Baseline (5 min)

Show data in plaintext, query patterns:

// Insert plaintext
db.patients.insertOne({
  ssn: "123-45-6789",
  dob: new Date("1980-05-15"),
  diagnosis: "Hypertension"
});

// Query plaintext
db.patients.findOne({ ssn: "123-45-6789" });

Demonstrate: Data visible in MongoDB Compass, logs, backups

Phase 2: Enable QE (10 min)

1. Create data encryption keys
2. Define encrypted fields schema
3. Create encrypted client
4. Insert encrypted data

// Create DEK
const keyId = await clientEncryption.createDataKey("local");

// Define schema
const schema = {
  "patients": {
    encryptedFields: {
      fields: [
        { path: "ssn", keyId, bsonType: "string", queries: { queryType: "equality" } },
        { path: "dob", keyId, bsonType: "date", queries: { queryType: "range", min: ..., max: ... } },
        { path: "diagnosis", keyId, bsonType: "string" }  // Random
      ]
    }
  }
};

// Insert encrypted
const encryptedClient = new MongoClient(uri, { autoEncryption: { keyVaultNamespace, kmsProviders, encryptedFieldsMap: schema } });
await encryptedClient.db("hospital").collection("patients").insertOne({
  ssn: "123-45-6789",  // Encrypted deterministic
  dob: new Date("1980-05-15"),  // Encrypted range
  diagnosis: "Hypertension"  // Encrypted random
});

Phase 3: Query Encrypted Data (10 min)

Show what queries work:

// ✅ Works: Equality on deterministic field
await patients.findOne({ ssn: "123-45-6789" });

// ✅ Works: Range on range field
await patients.find({ dob: { $gte: new Date("1980-01-01") } });

// ❌ Fails: Query on random field
await patients.findOne({ diagnosis: "Hypertension" });  // No results (encrypted)

Show in Compass: Ciphertext blobs, metadata collection

Phase 4: Compliance Mapping (5 min)

Connect to compliance requirements:

• "This meets HIPAA §164.312(a)(2)(iv) encryption requirement"
• "Key vault in AWS KMS provides FIPS 140-2 compliance"
• "Access controls via MongoDB RBAC satisfy HIPAA §164.308(a)(4)"

Python Tool Details

1. Vertical Analyzer

Input: Vertical name

Output: Sensitivity analysis

{
  "vertical": "healthcare",
  "sensitive_fields": [
    { "name": "ssn", "type": "deterministic", "regulation": "HIPAA", "reason": "PHI identifier" },
    { "name": "dob", "type": "range", "regulation": "HIPAA", "reason": "PHI demographic" },
    { "name": "diagnosis", "type": "random", "regulation": "HIPAA", "reason": "PHI medical" }
  ],
  "plaintext_fields": ["fullName", "city"],
  "compliance_frameworks": ["HIPAA", "HITECH"]
}

2. Demo Generator

Input: Vertical analysis

Output: Complete demo code (Node.js + Python)

• setup.js — Create keys, schema
• insert.js — Insert encrypted data
• query.js — Query examples
• README.md — Workshop guide

3. Compliance Mapper

Input: Vertical + framework

Output: Compliance mapping document

# HIPAA Compliance Mapping

## §164.312(a)(2)(iv): Encryption and Decryption
✅ MongoDB Queryable Encryption provides encryption at rest
✅ TLS 1.2+ provides encryption in transit

## §164.308(a)(4): Access Controls
✅ MongoDB RBAC restricts key vault access
✅ AWS KMS provides key access audit trails
...

Common Patterns

Pattern 1: Healthcare Patient Lookup

Scenario: Find patient by SSN, show non-PHI

// Encrypted: ssn, dob, diagnosis
// Plaintext: fullName, city (for UX)

const patient = await patients.findOne({ ssn: "123-45-6789" });
// Returns: { fullName: "John Doe", city: "Boston", ssn: <encrypted>, ... }

Pattern 2: Finance Transaction Search

Scenario: Lookup by last 4 digits, hide full card

// Encrypted: cardLast4 (deterministic), cardNumberFull (random)
// Plaintext: merchantName, amount

const transactions = await txns.find({ cardLast4: "4242" });
// Returns masked card, plaintext merchant/amount

Pattern 3: Government Clearance Query

Scenario: Find citizens with clearance level >= 3

// Encrypted: ssn (deterministic), clearanceLevel (range)
// Plaintext: state, registrationDate

const citizens = await citizens.find({ clearanceLevel: { $gte: 3 } });

Workshop Agenda (60 min)

0-10 min: Introduction

• Why encryption? (breaches, compliance)
• Encryption at rest vs in use vs QE
• Demo preview

10-25 min: QE Deep Dive

• Three encryption types (deterministic, range, random)
• Key management (AWS KMS, local)
• Schema definition

25-45 min: Live Demo

• Phase 1: Plaintext baseline
• Phase 2: Enable QE
• Phase 3: Query encrypted data
• Show in Compass (ciphertext)

45-55 min: Compliance Mapping

• Map to HIPAA/PCI-DSS/FISMA
• Production considerations
• Key rotation, access controls

55-60 min: Q&A + Resources

Compliance Cheat Sheet

Production Considerations

1. Performance

Impact:

• Encrypted writes: ~10-20% slower (encryption overhead)
• Encrypted reads: ~5-10% slower (decryption)
• Range queries: Higher overhead than equality

Mitigation:

• Index plaintext fields for filtering
• Use deterministic sparingly (lower security)
• Cache decrypted results when safe

2. Key Rotation

Manual process:

1. Create new data encryption key
2. Re-encrypt data with new key
3. Update schema to reference new key

Automation: Use scripts, schedule quarterly

3. Backup & Disaster Recovery

Key vault backup:

• AWS KMS: Automatic replication
• Local keys: Backup to secure storage

Encrypted data backup:

• Backup encrypted documents (safe)
• Restore requires key vault access

Quality Checklist

Before workshop:

• Vertical analysis complete
• Demo code tested locally
• Compliance mapping reviewed
• AWS KMS credentials configured (or local key ready)
• MongoDB Compass installed (show ciphertext)
• Timing rehearsed (60 min)

During demo:

• Show plaintext first (contrast)
• Explain each encryption type clearly
• Demo all three query patterns
• Show Compass (visual proof)
• Map to compliance requirements
• Leave time for Q&A

When to Use vs. Other Tools

References

• QE Compliance Guide: references/qe-compliance-guide.md
• Industry Patterns: references/industry-patterns.md
• Official docs: https://www.mongodb.com/docs/manual/core/queryable-encryption/
• Compliance: https://www.mongodb.com/compliance

Credits

Michael Lynn — mlynn.org · @mlynn · LinkedIn · GitHub

Next steps after workshop:

1. Customer-specific schema design
2. Production KMS setup
3. Performance testing
4. Compliance audit preparation