For county jails
Every inmate message, screened immediately.
AI reads your live message feed and flags threats, contraband and fraud for your staff in seconds.
What it catches
- Threats
- Contraband
- Escape plans
- Weapons
- Card and bank numbers
- Coded language
One queue, ranked by risk.
Category, confidence and the reason. Escalate or clear in one click.
Try the demo queueReview queue
Screened this week · 6,184 messages · 22 flagged
- Sensitive identifiers97%
Full card number with expiry, CVV and PIN.
DC-40533 · B-Pod
- Escape95%
Arranges an outside vehicle during a scheduled medical transport and notes escort size.
DC-40502 · Medical
- Threats / extortion94%
Implied harm toward a named person linked to cooperating with police, tied to a court date (possible witness intimidation).
DC-40117 · B-Pod
- Contraband / criminal activity89%
Instructions to soak paper with a substance and mail it under another sender's name (drug-soaked paper).
DC-40177 · B-Pod
Tested on real inmate messages
100%
of threats caught
in testing
93%
of card and bank numbers caught
in testing
0.2%
false-flag rate on threats
random sample
Tested on real messages from an archive of nearly 10 million written by incarcerated people.
How we measured it
A sample of 1,298 real messages written by people in federal facilities, labelled by two independent AI reviewers. Human-labelled validation is in progress. Your pilot measures accuracy on your own traffic.
Threats / extortion
- Recall
- 100%
- 95% CI 80.6 to 100%
- Precision
- 53%
- Threshold
- 0.6
- False-flag rate
- 0.2%
Sensitive identifiers
- Recall
- 93%
- 95% CI 82.4 to 97.1%
- Precision
- 74%
- Threshold
- 0.3
- False-flag rate
- 0.3%
Contraband / criminal activity
In calibration
Other categories
In calibration
| Category | Recall | Precision | False-flag rate | Threshold |
|---|---|---|---|---|
| Threats / extortion | 100%95% CI 80.6 to 100% | 53% | 0.2% | 0.6 |
| Sensitive identifiers | 93%95% CI 82.4 to 97.1% | 74% | 0.3% | 0.3 |
| Contraband / criminal activity | In calibrationMeasured during your pilot and published once calibrated. | |||
| Other categories | In calibrationMeasured during your pilot and published once calibrated. | |||
Security controls for the pilot
Controls modeled on the FBI CJIS Security Policy. Formal CJIS work starts when you sign.
- US-hosted, encrypted
- Named users with MFA
- Full audit log
- Attorney contacts excluded
- No training on your data
- Your retention rules
Free 90-day pilot
Judge it on your own traffic.
- No cost, no purchase contract
- Runs alongside your current process
- Accuracy readout at day 90
Built by Jordan Rejaud, founder. Every request comes straight to him at admin@readwarden.com.
- Does it block messages?
- No. It flags. Your staff make every decision.
- What about attorney mail?
- Privileged contacts on your list are never screened or stored.
- What happens after 90 days?
- You see the numbers and decide. No auto-renewal.