A nighttime homicide in a remote agricultural field. The victim was recovered the next morning. By the time the case reached us, there was a body, a scene, and not much else — the kind of file where the conventional trail is cold before anyone's had coffee.
So we went back to the scene. Not physically — in the data.
Geofence resolution at the location, during the window the offense had occurred, surfaced a single anomalous device in close proximity to the body. We traced it to its overnight residence and a corresponding IP address. One device on the record, where it had no innocent reason to be, surfacing both an identity and a location. The scene had been quiet for months. The record of who was standing in it never went anywhere.
That's the capability most investigators don't realize they have: a scene can be revisited after the fact, and the devices that were present can be surfaced — sometimes the suspect, often the witnesses, and frequently both.
The state stops looking. You don't have to.
Here's the thing about a police investigation: it's built to support a charge. Once detectives have a suspect they like, the lead development slows and then stops. The file gets closed around a theory. Everyone who doesn't fit that theory — the witness no one canvassed, the device that doesn't belong to the defendant — simply never makes it into the record.
For a defense investigator, that gap is the whole opportunity. When you can independently geofence the scene and surface every device that was present during the relevant window, you're building something the prosecution's file does not contain: an unfiltered picture of who was actually there. A witness the police never identified. A device that places someone other than your client at the scene. Corroboration — or contradiction — of the state's timeline.
We had a separate matter, a shooting at a regional facility, where geofence resolution during the offense window surfaced eleven distinct devices with witness-consistent presence, plus one anomalous profile that didn't fit the others. All twelve resolved to identifiable subjects. Eleven of them were people no one had interviewed — and any one of them could have seen the thing that changes a case.
Every device surfaced is work that didn't exist before
This is where it matters to your practice, not just your client's case.
A defense engagement usually starts narrow. Review the file. Run down a name. Maybe re-interview the one witness in the discovery. Then you geofence the scene, and eleven new people land on your desk — eleven witnesses who have to be located, backgrounded, approached, interviewed, and written up. A device that doesn't fit the state's theory turns into an alternate-suspect thread you can actually pull.
None of that work existed as a lead an hour earlier. It's not padding — it's legitimate, case-advancing investigation that materially serves the client, generated by a capability the rest of the field doesn't have. One geofence can turn a one-week file review into a multi-witness investigation. The hours follow the value, and the value is real: you're handing the defense avenues the prosecution never built.
And it changes what kind of work comes to you. An investigator who can break a stalled criminal case open — who can walk into a defense attorney's office with a witness list the state never had — is the investigator that attorney calls on the next hard case, and the one they refer to their colleagues. Case-cracking capability is how you move up the value chain in criminal defense work.
What it surfaces — and what it doesn't
I'll hold to the same honesty here that I'd defend on a stand, because in criminal work the stakes don't allow anything less.
Geofencing a scene surfaces leads, not verdicts. A device places a device at or near a location with high confidence — not, by itself, a named person, and certainly not a guilty one. Presence at a scene is not participation in a crime; plenty of the devices you surface belong to people who were simply nearby. Every name that comes off a geofence still has to be located, identified, and interviewed before it's worth anything. The capability hands you the thread. You still have to follow it.
Each finding carries its calibration — what the signal proves, what the pattern infers, and what stays open pending the records that would confirm it. That discipline is exactly what keeps a surfaced witness or an alternate-suspect lead useful instead of a liability when it's challenged. And, as with all of our work, scene resolution runs on a documented predicate — defense engagement, attorney direction, or law-enforcement support.
But the next time a criminal or civil file lands on your desk looking cold — a scene, a disputed account, and a client who's sure the truth is somewhere no one's looked — remember that the scene isn't as empty as the file makes it look. Everyone who was standing in it left a record. The question is whether anyone's gone back to read it.