CASE STUDY
How Aurora Public Defense Office Used AI transcription to win a case
The officer had discussed holding my client at gunpoint prior to making contact. I couldn’t hear it but computer transcription picked it up.
Elizabeth D. Cadiz, Chief Deputy Public Defender, Aurora CO
THE CHALLENGE
The rise in audiovisual evidence
With the growing volume of bodycam footage and police interview video being introduced as evidence in criminal cases, the Aurora Public Defender’s Office needed a better way to manage and go through all of it.
Prosecutors would often introduce selective parts of the footage that supported their case, leaving out context that could have favored the defense. As a result, their whole team was overwhelmed by the endless sifting of hours of footage to identify inconsistencies, contradictions, or exculpatory evidence.
Moreover, the body-worn camera footage came in fragmented formats, on flash drives and discs, without a centralized or easily accessible system, making it even more difficult to manage the increasing hours of audiovisual evidence.
THE SHIFT
Reduct.Video implementation and impact
Reduct has become an essential tool for public defenders in Aurora to triage, search for key moments, and clip out footage to present as evidence in court.
In one case, their Chief Deputy Public Defender Elizabeth’s primary argument centered around racial discrimination against her African American client. Using Reduct’s search features, she was able to find a particular racially inflammatory phrase used by a witness. This became a key piece of evidence in the case centered around race, and her client was found not guilty.
In another case, police bodycam footage captured a faint audio of two officers talking- something she hadn’t initially noticed. But Reduct’s AI transcription picked it up, she was then able to build a case around that evidence, ultimately securing a not-guilty verdict.
Key Reduct.Video features in use at Aurora
1. AI TRANSCRIPTION
Reviewing discovery with time-stamped transcripts
Before using Reduct, Aurora Public Defenders had to manually identify the exact second or millisecond and follow a complex workflow to create video exhibits from body-worn camera footage. With Reduct, they can now select the relevant transcript, highlight it, and download just the highlighted section of a lengthy bodycam footage with just a few clicks.
With Reduct, they get access to accurate AI transcriptions within minutes of uploading their audio and video discovery. Every word in the transcript is synced to its exact moment in the video. Instead of flipping through pages, the public defenders and their assistants can instantly click on a word in the transcript, and easily jump to that exact moment in the video, speeding up review.
2. SEARCH
Finding key phrases to build a defense
When developing a theory of defense, identifying specific phrases used in body-worn camera footage can be critical. Before Reduct, Aurora Public Defenders didn’t have an efficient way to search within video discovery, leaving defenders to review hours of footage manually.
With Reduct’s transcription and search functionality, they can now quickly search for phrases that support their theory of defense and locate key statements that strengthen their case.
3. REELS
Extracting key moments for video exhibits
Before using Reduct, Aurora Public Defenders had to manually identify the exact second or millisecond, and had a complex workflow to create video exhibits from body-worn camera footage. With Reduct, they can now select the transcript to highlight and download just the highlighted section from a lengthy bodycam with a few clicks.
They have shifted from manually time-stamping a paper transcript to clipping out the exact moments on the fly.
By leveraging Artificial Intelligence to alleviate the painpoint of too-much evidence and not-enough people hours, Reduct lets defense lawyers focus their energies on defending the public.
