Why Single-Call APIs Break at Enterprise Scale
A legal firm reviewing 2,000 deposition excerpts. A bank screening 50,000 loan applications. A healthcare system summarizing 10,000 patient intake forms. If your anonymization API processes one document per call, you're looking at:
- Latency — 10,000 sequential calls × 200ms each = 33 minutes minimum
- Rate limits — burst limits kill parallelism beyond 50–100 concurrent calls
- Cost — per-request overheads (connection, auth, TLS handshake) add up at scale
- Complexity — managing 10,000 async calls, retries, and partial failures in application code
cloak.business's cloak_batch_analyze tool was built specifically for this pattern.
Batch Analyze — 1 to 100 Texts Per Call
cloak_batch_analyze accepts an array of text items and returns entity detection results for all of them in a single round-trip. Each item in the batch can have its own language hint, entity filter, and method override — the batch isn't just "call 100 times internally." It's a single vectorized pass over all inputs.
{
"tool": "cloak_batch_analyze",
"items": [
{
"id": "doc-001",
"text": "John Schmidt, DOB 1981-04-12, applied for loan #882244",
"language": "de"
},
{
"id": "doc-002",
"text": "Customer Maria Lopez, IBAN ES79 2100 0813 6101 2345 6789",
"language": "es"
}
// ... up to 100 items
],
"method": "replace"
}
// Response:
{
"results": [
{
"id": "doc-001",
"anonymized_text": "[PERSON_1], DOB [DATE_1], applied for loan #[ID_1]",
"entities_found": 3
},
{
"id": "doc-002",
"anonymized_text": "Customer [PERSON_1], IBAN [IBAN_1]",
"entities_found": 2
}
]
}
Chrome Extension — Browser-Level Coverage for the Whole Team
Enterprise AI risk isn't just in automated pipelines. It's in every employee who copies client data into ChatGPT, pastes a contract clause into Claude, or uploads a financial summary to Gemini. The API protects your pipelines. The Chrome Extension protects your people.
What the Chrome Extension Does
- Intercepts text in ChatGPT, Claude, Gemini, DeepSeek, Perplexity, and Copilot before submission
- Runs detection against cloak.business's API — 320+ entity types, 70+ countries
- Replaces PII with deterministic tokens in real time, transparently
- De-anonymizes AI responses so employees see original values — round-trip deanonymization
- Works without requiring employees to change their workflow
The extension is deployable via Chrome Enterprise policy, so IT can push it to every browser in the org without individual user installation. Per-user keys can be pre-configured via managed policy, tied to employee accounts, and revoked centrally.
Target Use Cases and Industries
Financial Services
Banks and insurance companies run nightly batch jobs that classify, summarize, or extract data from thousands of customer communications. cloak.business integrates into these pipelines at the ingest stage — every document is anonymized before it touches any AI model. The batch API makes this economical even at millions of records per day.
Legal and Professional Services
Law firms use AI for contract review, discovery review, and research. Paralegals use ChatGPT daily. The Chrome Extension covers individual AI usage; the MCP Server covers Claude-assisted contract analysis; and the batch API covers bulk review pipelines — three deployment points, one platform, one subscription.
Healthcare and Life Sciences
Patient intake forms, clinical trial data, radiology reports. HIPAA and GDPR require that PHI not reach third-party AI providers. cloak.business's batch API processes clinical documents in bulk; image OCR handles scanned forms; and the Chrome Extension catches any manual AI queries from clinical staff.
HR and Recruiting
Recruitment teams use AI to screen CVs and draft job offers. Each CV is rich with PII: full name, address, phone, email, employment history. cloak.business's batch API anonymizes incoming CV batches before they reach any AI screening model — and the Chrome Extension protects individual recruiters using ChatGPT to draft rejection letters.
Pricing — Built for Teams, Not Just Developers
cloak.business pricing reflects enterprise realities — volume, teams, and compliance requirements — rather than the per-seat developer model.
| Plan | Batch API | Image OCR | Chrome Extension |
|---|---|---|---|
| Free | — | 50 credits | — |
| Pro | Up to 50/call | 500 credits/mo | 1 seat |
| Business | Up to 100/call | Unlimited | Team seats |
| Enterprise | Custom | Custom | Managed policy |
See cloak.business/pricing for current pricing details. All plans start with a 14-day free trial of Business features.
cloak.business vs anonym.legal — When to Choose Which
| You need... | cloak.business | anonym.legal |
|---|---|---|
| Process 100 docs at once | Yes — batch_analyze | 1 doc per call |
| Protect team browser AI usage | Yes — Chrome Extension | MCP Server only |
| Anonymize scanned images | Yes — image OCR tools | Text only |
| Simple developer MCP setup | Yes (10 tools) | Yes — 7 tools, simpler |
| Cover 70+ countries | Yes — 320+ entities | 48 languages |
| Round-trip deanonymization | Yes | Yes |
If you're an individual developer or small team working with text pipelines, anonym.legal is the simpler choice. If you're an enterprise protecting a team, processing images, or running bulk pipelines, cloak.business covers the full stack.
Both products share the same infrastructure: Microsoft Presidio detection engine, Hetzner servers in Germany, ISO 27001:2022 certification, and deterministic anonymization with reproducible token generation. The difference is in scale, capabilities, and deployment model — not in the underlying security architecture.