Built for healthtech billing teams

Redact PHI from voice call logs
before they hit your LLM

Shroud scrubs MRNs, amounts, and patient identifiers from billing narratives at wire speed. One API call, 28 detectors, sub-15 ms p95.

Try your own data Book a demo
YOUR BILLING NARRATIVE
CLEANED FOR YOUR LLM
Output will appear here...
Why LunaBill needs Shroud

Billing narratives are PHI landmines

PHI in every transcript

Your AI voice callers handle patient names, MRNs, and billing amounts in every call. Shroud scrubs them before they hit your LLM training data.

Scaling past $764K ARR

As you grow, manual redaction breaks. Shroud gives you an API that handles 5M characters/month on the Startup plan — enough for thousands of calls.

Zero-integration for Python

Drop Shroud into your voice pipeline with three lines of Python. No schema changes, no new infrastructure — just wrap your logging call.

Your integration

Three lines in your Python voice stack

# pip install shroud
from shroud import Client

shroud = Client(api_key="sk_live_...")

# In your voice call handler:
call_transcript = "User MRN 8471293 paid $42.50 with card ending 4111"
# ↓ wrap your log line
cleaned = shroud.redact(call_transcript)
log.info("Call completed: %s" % cleaned)