- Days
- Hours
- Minutes
The most accurate and intelligent AI in public safety
TRULEO's models
Speech-to-text Transcription
We use a proprietary, internally-built model to transcribe BWC audio to text with industry-leading accuracy that exceeds humans. Our model is purpose built in that it focuses on the interaction between the officer wearing the camera and with whom they are conversing with, ignoring other parties that may be much further away from the interaction.
Speaker Diarization
Speaker diarization is the task of a machine deciding "who said what?" We use industry-leading speaker diarization that anonymously separates out community member audio from officer audio, then specifically identifies the officer wearing the camera using an Officer ID model.
Translation
Our model automatically detects and transcribes Spanish, then translates the text to English.
Event Recognition
Our Large Language Models (LLM) analyze the transcript and identify dozens of events with 99% accuracy. See a full list of event label definitions here.
Generative AI
We use industry-leading models from AWS Bedrock and OpenAI in secure environments for generative AI tasks such as enhanced narrative creation, summarization, and video insights.
A secure and compliant way to implement AI
We maintain strict security and chain of custody over your data
No Video Storage:
We stream data from evidence APIs and only store the analysis in our databases. When a user plays back a video, it proxies a new streaming connection to your evidence platform. None of your videos are ever retained.
Limited Retention of Analytics Data:
Regardless of the evidence preservation period setting for any BWC video in the department's BWC video management system, Our application's data retention policy is set to sixty (60) days, after which time all data stored is automatically deleted.
No processing of Restricted Content:
Any BWC video that is marked restricted (or "locked down" etc.) or is removed/deleted from the department's BWC video management system for any reason will not be analyzed or be accessible.
CJIS Compliant:
We are CJIS Compliant and deployed entirely in AWS GovCloud. All data is encrypted in transit and at rest.
PII Redaction:
TRULEO's LLMs also automatically redact personal identifiable information (PII) to protect community member privacy.
De-escalation attempts
Community gratitude
Upset person
High composure
How it works
Responsible transcription
Truleo's Responsible Transcription automatically redacts personal identifiable information and separates officer speech.
Cutting-edge Natural Language Processing
A department's internal team reviews highly professional clips for potential highlight reels.
Truleo detects. Human confirms.
Sometimes, context really matters. Any events that require a second set of eyes are surfaced to supevisors to confirm what actually happened.
Conflict-Free Government Purchasing
We are not a weapons manufacturer.
We do NOT sell body‑worn cameras.
We have a strong commitment to protecting officer and civilian privacy.
We are 100% MADE IN AMERICA with 100% onshore development.
Data management and storage
Security FAQs
What security frameworks do you follow?
TRULEO complies with CJIS Security Policy and FedRAMP standards, leveraging AWS GovCloud infrastructure.
Does TRULEO support role-based access control (RBAC)?
Yes, TRULEO provides RBAC capabilities, restricting access based on user roles.
What authentication mechanisms are in place?
TRULEO supports Single Sign-On (SSO) with customer identity providers and requires Multi-Factor Authentication (MFA).
How is sensitive information like PII and CJIS data handled?
Sensitive data is encrypted using FIPS 140-2 standards at rest and in transit. Automated redaction models are used to remove sensitive details from body-worn camera transcripts.
Where is customer data stored?
Data is stored in AWS GovCloud, segmented within encrypted AWS Aurora RDS instances.