Effortlessly Redact Personal Data from Call and Chat Transcripts

The Problem:

Traditional Transcript Redaction is a Lot of Work

Call and chat transcripts are notoriously messy and can be some of the hardest data to redact properly. 

Current solutions aren’t cutting it. 

Human redaction is slow, expensive, and notoriously inaccurate. Automated systems, like regexes, are faster and cheaper, but unable to process idiosyncratic data, like the back-and-forth of a call transcript or the many typos within transcripts. And building your own AI system takes so much time, money, and effort it’s almost certainly a better choice to buy a proven solution.

Enter Private AI:

Fast, Accurate, and Secure Redaction Software

Private AI leverages the latest advancements in transformer architectures to pick out personal data (PII, PHI, PCI, etc.) from transcripts with incredibly high accuracy: 99.5%+*. Test it out for yourself with our web demo.

No regexes, no dictionaries, no rule-based systems.

The system is optimized for the idiosyncrasies of transcripts and chat logs, such as disfluencies, emojis, and internet slang, and can support regexes provided by customers upon request.

Private AI provides the redaction capabilities for a number of leading:

Automatic Speech Recognition (ASR) Transcription Providers

Contact Center Transcription Services

Chatbot Platforms

Telehealth Platforms

Why Private AI

Unrivalled Accuracy

Private AI uses the latest advancements in machine learning to achieve remarkable accuracy out of the box. See how we stack up against our competitors in our technical whitepaper

Private AI
Major Cloud Provider 2
Open Source Software 2
Open Source Software 1
Major Cloud Provider 1
Major Cloud Provider 3
0.80 0.90 1

Try it yourself on your own data:

We provide a speech-to-text transcription API and needed to bring our redaction of credit cards, SSNs, and other personal financial and health information up to the highest accuracy level possible. Private AI made that quick and easy – now our accuracy numbers are through the roof and our customers are happy, which has been amazing. Plus they were remarkably easy to integrate into our existing workflows, which saved us a lot of time and effort compared to building something in-house.

Dylan Fox
CEO, Assembly AI

99.5%+ Accuracy

Number quoted is the number of PII words missed as a fraction of total number of words. Computed on a 268 thousand word internal test dataset, comprising data from over 50 different sources, including web scrapes, emails and ASR transcripts.

Please contact us for a copy of the code used to compute these metrics, try it yourself here, or download our whitepaper.


Tested on a dataset composed of messy conversational data containing sensitive health information. Download our whitepaper for further details, as well as how we perform on precision and F1-score or contact us to get a copy of the evaluation code.


Testé sur un ensemble de données composé de données conversationnelles désordonnées contenant des informations de santé sensibles. Téléchargez notre livre blanc pour plus de détails, ainsi que nos performances en termes d’exactitude et de score F1, ou contactez-nous pour obtenir une copie du code d’évaluation.