Announcing Incident Source and Project Radiant Glow — AI-Powered Real-Time Incident Intelligence

[blantonl]

Dallas–Fort Worth, TX — Incident Source, a new offering part of the RadioReference Network, is excited to unveil Radiant Glow, a next-generation AI intelligence platform that transforms live radio communications into real-time incident and event awareness.

IncidentSource - Radiant Glow - Incident Intelligence

Radiant Glow listens to continuous streams of public safety and operational radio traffic from Broadcastify Calls, applies advanced artificial intelligence, and surfaces meaningful incidents as they unfold. Instead of raw audio or fragmented reports, users gain a clear, correlated view of events: what’s happening, where it’s happening, and how situations evolve over time.

Built on Broadcastify's proven radio data infrastructure and the scale of the RadioReference Network, Radiant Glow introduces a new layer of intelligence:

• Real-time signal ingestion from radio communications traffic
• AI-driven detection and classification of incidents and events
• Automated correlation of related transmissions into unified incident threads

Radiant Glow is currently operating in technology preview across the Dallas–Fort Worth Metroplex, where it is actively correlating live signals into structured, actionable incident intelligence. Expansion to additional regions is planned as the platform evolves.

Lindsay Blanton, CEO of RadioReference.com, shared his perspective on the launch:

“Radiant Glow represents the most powerful AI-driven public safety intelligence platform currently available to the general public. We’re excited to introduce this technology preview and give users an early look at what’s possible when advanced AI is applied to radio communications at scale.

This project just the beginning. We look forward to expanding Radiant Glow into additional locations, introducing new features and functionality, and launching even more projects built on this platform in the months ahead.”
----------------------------------------------------------------------------------------------------------------
Updates 1/3/2026

Will there be an API?

Hell yes.

"A lot of the transcriptions are not working well, and there are false incidents being created, why?"

We're still tuning the transcriptions model and escalations process, but the good news is that we're not actually sending out alerts or notifications just yet. The purpose of this platform is to be an intelligence gathering platform for those who need incident alerting and notifications (other platforms, or maybe even us). The other good news is the platform is setup to self learn over time, so things will only get better over time as we progress and teach the model.

"Is there a way we can notify you about inaccuracies?"

Yes, we should have the ability to report problems with incidents generated shortly, including rollback capabilities, reasons why ("Hey! This is a training events etc"). Stay tuned.

"When are you going to cover my area"

The basis for this entire platform is Broadcastify Calls coverage. If your area has excellent Broadcastify Calls coverage, it's a candidate for this platform. You can send me a message to support@broadcastify.com giving me a heads up on a "context" that you'd like to see added and I'll consider it. A "context" in the sense of this platform is a grouping of talkgroups/frequencies that we correlate incidents across. This typically right now is an agencies Fire Dispatch and Fire Tactical groups.

This should also be a nudge to those who want to see their area covered, if there isn't broadcastify calls coverage, we need it, so adding calls nodes to fill out coverage will mean incident source will cover your area.

Also, in the past two days we've added the following contexts/locations

Boston Fire
South Metro Fire/Rescue (Denver Area)
West Metro Fire/Rescue (Denver Area)
Central Fire/Rescue (Denver Area)
San Diego Fire
San Diego County Fire
New Orleans Fire/EMS
Jefferson Parish Fire
San Antonio Fire/EMS
Bexar County Fire Alarm
Houston Fire/EMS
Austin Fire/EMS
Travis County Fire/EMS
Atlanta Fire
Chicago Fire
San Bernardino County CONFIRE
Philadelphia Fire

 

[Enforcer52]

Sign Me Up!

 

[Harold]

Certainly makes it easier to find a particular radio transmission, when looking for an address or something.

 

[wbloss]

The concept is great - there is alway some difficulty "hearing" the correct words: "Engine" becomes "Ninja" and "Eng" (short for ENgine) becomes "Edge". Maybe AI will get to know the mumbles better after a while.

 

[BenScan]

Neat tech, but I honestly think it will be another argument for some departments to encrypt. Looking at and listening to one of the calls it appears to need some training on what is training. Not all fire departments use dedicated training channels. This could lead to some false alerting.

 

[Elman88]

defiantly, sounds like a new begin sing me up !!!

 

[blantonl]

Neat tech, but I honestly think it will be another argument for some departments to encrypt. Looking at and listening to one of the calls it appears to need some training on what is training. Not all fire departments use dedicated training channels. This could lead to some false alerting.

We're working on training what is training. 😏

We're not sending any alerts out right now, instead we're using it as an intelligence platform to eventually supplement alerting and notification systems. Anyone with knowledge of these departments could have switched those transcripts into tactical mode (just viewing the transcripts) and seen that it was training and made the human determination right there, and you can quickly back it up by listening to the actual calls.

We'll have the system setup shortly where you could have notified us immediately about something listed there (I'm assuming some of those Ellis County Fires) as training incidents and we immediately could evaluate and roll back.

 

[DaveNF2G]

The concept is great - there is alway some difficulty "hearing" the correct words: "Engine" becomes "Ninja" and "Eng" (short for ENgine) becomes "Edge". Maybe AI will get to know the mumbles better after a while.

Poor articulation seems to be a growing problem. I believe it tracks with reliance on printouts and text paging.

 

[blantonl]

Poor articulation seems to be a growing problem. I believe it tracks with reliance on printouts and text paging.

 

[slatescan]

I dont know if you noticed or not but the Philly fire channels are showing up as San Bernardino County, loving this so far Lindsay!

 

[blantonl]

Some updates and answers to questions we've received

Will there be an API?

Hell yes.

"A lot of the transcriptions are not working well, and there are false incidents being created, why?"

We're still tuning the transcriptions model and escalations process, but the good news is that we're not actually sending out alerts or notifications just yet. The purpose of this platform is to be an intelligence gathering platform for those who need incident alerting and notifications (other platforms, or maybe even us)

"Is there a way we can notify you about inaccuracies?"

Yes, we should have the ability to report problems with incidents generated shortly, including rollback capabilities etc. Stay tuned.

"When are you going to cover my area"

The basis for this entire platform is Broadcastify Calls coverage. If your area has excellent Broadcastify Calls coverage, it's a candidate for this platform. You can send me a message to support@broadcastify.com giving me a heads up on a "context" that you'd like to see added and I'll consider it. A "context" in the sense of this platform is a grouping of talkgroups/frequencies that we correlate incidents across. This typically right now is an agencies Fire Dispatch and Fire Tactical groups.

This should also be a nudge to those who want to see their area covered, if there isn't broadcastify calls coverage, we need it, so adding calls nodes to fill out coverage will mean incident source will cover your area.

Also, in the past two days we've added the following contexts/locations

Boston Fire
South Metro Fire/Rescue (Denver Area)
West Metro Fire/Rescue (Denver Area)
Central Fire/Rescue (Denver Area)
San Diego Fire
San Diego County Fire
New Orleans Fire/EMS
Jefferson Parish Fire
San Antonio Fire/EMS
Bexar County Fire Alarm
Houston Fire/EMS
Austin Fire/EMS
Travis County Fire/EMS
Atlanta Fire
Chicago Fire
San Bernardino County CONFIRE
Philadelphia Fire

 

[blantonl]

I dont know if you noticed or not but the Philly fire channels are showing up as San Bernardino County, loving this so far Lindsay!

Whoopies. Fixed

 

[d119]

Neat tech, but I honestly think it will be another argument for some departments to encrypt.

At the risk of my account (as a regular contributor to lowband logs), notice you don't see any Orange County, CA fire stuff on the list. This sure doesn't help the cause for them to un-encrypt. I'll do a full stop there.

 

[Enforcer52]

Is this going to be for Fire/EMS only, or is this the area of concentration at the beginning?

 

[blantonl]

Is this going to be for Fire/EMS only, or is this the area of concentration at the beginning?

Yes, Fire EMS is the primary concentration right now. Because we can also encompass the vast majority of major law enforcement events from Fire/EMS sources. I'm undecided right now on actually tracking medical calls at this time because there isn't going to be much of a reason to do so based on the objectives of the project. I'm leaning towards removing and filtering them out completely

 

[Enforcer52]

I agree with filtering out medical calls, way too many very, very minor reasons folks call for EMS.

 

[d119]

@blantonl, I could see insurance companies using aggregated medical call data broadcast over the air (addresses, etc) for nefarious purposes. Something to consider. Granted they see the bills anyway, but if a provider were able to determine "high risk" based on location/call type, well.. I'm sure they can get that information elsewhere, but it doesn't help to make it easier. I think your filtering of medical calls is wise and right on the money (pardon the pun).

 

[blantonl]

@blantonl, I could see insurance companies using aggregated medical call data broadcast over the air (addresses, etc) for nefarious purposes. Something to consider. Granted they see the bills anyway, but if a provider were able to determine "high risk" based on location/call type, well.. I'm sure they can get that information elsewhere, but it doesn't help to make it easier. I think your filtering of medical calls is wise and right on the money (pardon the pun).

I hate to tell you, but there are already NUMEROUS companies doing what you're seeing now on Incident Source, it's just not publicly available.

That's why the marketing messaging includes the phrase "Radiant Glow represents the most powerful AI-driven public safety intelligence platform currently available to the general public."

Before RG was released, there wasn't really any platform available to the general public like you see now.

I've gotten emails this week from numerous people and professionals seeing something like this in action for the first time, and the general consensus is "holy sh*t this is amazing"

 

[Razorback55]

30 years ago, a professional with a high level of clearance told me that the NSA did not have people to listen to intercepted communications and make reports, that it was done automatically by computers and that the computer only made reports on interesting topics, everything else was simply archived.

I didn't believe it at the time...

In the end, it was true. But now that we have AI and the NSA already had it 30 years ago, where are they today?

 

[AEFI]

Dallas–Fort Worth, TX — Incident Source, a new offering part of the RadioReference Network, is excited to unveil Radiant Glow, a next-generation AI intelligence platform that transforms live radio communications into real-time incident and event awareness.

IncidentSource - Radiant Glow - Incident Intelligence

View attachment 196142
Radiant Glow listens to continuous streams of public safety and operational radio traffic from Broadcastify Calls, applies advanced artificial intelligence, and surfaces meaningful incidents as they unfold. Instead of raw audio or fragmented reports, users gain a clear, correlated view of events: what’s happening, where it’s happening, and how situations evolve over time.

Built on Broadcastify's proven radio data infrastructure and the scale of the RadioReference Network, Radiant Glow introduces a new layer of intelligence:

• Real-time signal ingestion from radio communications traffic
• AI-driven detection and classification of incidents and events
• Automated correlation of related transmissions into unified incident threads

Radiant Glow is currently operating in technology preview across the Dallas–Fort Worth Metroplex, where it is actively correlating live signals into structured, actionable incident intelligence. Expansion to additional regions is planned as the platform evolves.

Lindsay Blanton, CEO of RadioReference.com, shared his perspective on the launch:

“Radiant Glow represents the most powerful AI-driven public safety intelligence platform currently available to the general public. We’re excited to introduce this technology preview and give users an early look at what’s possible when advanced AI is applied to radio communications at scale.

This project just the beginning. We look forward to expanding Radiant Glow into additional locations, introducing new features and functionality, and launching even more projects built on this platform in the months ahead.”
----------------------------------------------------------------------------------------------------------------
Updates 1/3/2026

Will there be an API?

Hell yes.

"A lot of the transcriptions are not working well, and there are false incidents being created, why?"

We're still tuning the transcriptions model and escalations process, but the good news is that we're not actually sending out alerts or notifications just yet. The purpose of this platform is to be an intelligence gathering platform for those who need incident alerting and notifications (other platforms, or maybe even us). The other good news is the platform is setup to self learn over time, so things will only get better over time as we progress and teach the model.

"Is there a way we can notify you about inaccuracies?"

Yes, we should have the ability to report problems with incidents generated shortly, including rollback capabilities, reasons why ("Hey! This is a training events etc"). Stay tuned.

"When are you going to cover my area"

The basis for this entire platform is Broadcastify Calls coverage. If your area has excellent Broadcastify Calls coverage, it's a candidate for this platform. You can send me a message to support@broadcastify.com giving me a heads up on a "context" that you'd like to see added and I'll consider it. A "context" in the sense of this platform is a grouping of talkgroups/frequencies that we correlate incidents across. This typically right now is an agencies Fire Dispatch and Fire Tactical groups.

This should also be a nudge to those who want to see their area covered, if there isn't broadcastify calls coverage, we need it, so adding calls nodes to fill out coverage will mean incident source will cover your area.

Also, in the past two days we've added the following contexts/locations

Boston Fire
South Metro Fire/Rescue (Denver Area)
West Metro Fire/Rescue (Denver Area)
Central Fire/Rescue (Denver Area)
San Diego Fire
San Diego County Fire
New Orleans Fire/EMS
Jefferson Parish Fire
San Antonio Fire/EMS
Bexar County Fire Alarm
Houston Fire/EMS
Austin Fire/EMS
Travis County Fire/EMS
Atlanta Fire
Chicago Fire
San Bernardino County CONFIRE
Philadelphia Fire

Yay SBco Commcenter (Is it my calls or someone else?)