SHERLOCK VIDEO ANALYSIS
12 evidence moments found
SENT THROUGH THE API
Recording → structured report
02:14
Scripted reading pattern
05:32
Synthetic voice signal
11:03
Deepfake risk signal
Each finding links to the exact moment in the recording.
APIs
Interview Recordings
Scripted Reading
Deepfake
HOW IT WORKS
SHERLOCK VIDEO ANALYSIS · CANDIDATE_INTERVIEW.MP4
Analysis complete
12 evidence moments found across video, audio, and response behavior.
02:14 · Scripted reading pattern
05:32 · Synthetic voice signal
08:47 · Repeated off-screen attention
11:03 · Deepfake risk signal
12:31 · Outside assistance detected
Evidence by timestamp
Every finding links back to the exact moment in the recording.
API response · Timestamped evidence · Human review
WHAT SHERLOCK CHECKS
Sherlock combines independent signals. No single signal decides the outcome.
Synthetic media
Deepfake video, cloned voice, lip-sync mismatch, and edited frames.
Scripted reading
Eye movement and response timing consistent with reading from a script.
Outside assistance
Coaching, second screens, and another person helping off camera.
Identity continuity
Changes in face, voice, or presence across the same recording.
AUTHENTICITY EVIDENCE, NOT ASSUMPTION
Turn deepfake interview risk into an evidence-backed hiring review.
When video or voice authenticity is in doubt, Sherlock AI turns deepfake detection signals into a reviewable evidence trail—helping recruiting and talent teams investigate synthetic-media risk, interview impersonation, and candidate fraud fairly.
REAL-TIME AUTHENTICITY REVIEW
Trace the media changes that point to manipulation.
Map face continuity, voice character, and frame texture across the call to isolate moments that may indicate generated or altered media.
Deepfake authenticity snapshot
ANALYZING
Aligned
Face continuity
Stable
Voice profile
Observed
Lip-sync trace
Ready
Reviewer evidence
Synthetic media signal pulse
Visual and voice consistency across the interview
AUTHENTICITY EVIDENCE
Build an audio-visual case, not a suspicion.
Correlate lip-sync drift, synthetic-voice cues, and image artifacts so reviewers can see which signals reinforce each other.
REVIEW QUEUE
Go straight to the frames and phrases that matter.
Timestamped media markers let reviewers inspect the exact visual or vocal change without replaying an entire interview.
Linked authenticity moments
02:14 · Facial continuity changed after join
08:47 · Lip-sync timing became inconsistent
12:31 · Voice profile drift detected
Open deepfake evidence trail →
Deepfake evidence handoff
Evidence package ready for review
Visual and voice signals, timestamps, and reviewer notes are organized in one place.
REVIEW READY
Hand reviewers a clear media-authenticity brief.
The final review package keeps linked clips, audio-visual signals, and context together for a defensible human decision.
BUILT FOR RECORDED INTERVIEWS
1
Upload
Video file or recorded interview
2
Analyze
Sherlock video-analysis API
3
Review
Timestamped fraud evidence
AI interview platforms
Analyze interviews recorded by automated screening and interview tools.
Human-led interviews
Review recordings from recruiter and hiring-manager calls.
Batch review
Submit many interviews and track each analysis in your workflow.
API-first workflow
Send recordings from your existing product and receive structured results.
THE SHERLOCK DIFFERENCE
From deepfake uncertainty to evidence-backed review.
Sherlock gives hiring teams the context to investigate synthetic-media concerns thoughtfully—before assumptions become decisions.
BEFORE SHERLOCK
A synthetic-media concern without context.
Reviewers are left piecing together visual and audio anomalies from memory, instinct, and incomplete notes.
• Unclear signals and subjective recall
• Slow, inconsistent investigation
• High-stakes decisions without proof
AFTER SHERLOCK
A clear authenticity trail to the next step.
Sherlock turns video and voice authenticity signals into linked evidence, giving every reviewer the context to assess a concern fairly.
✓ Timestamped, reviewable evidence
✓ Faster alignment across reviewers
✓ Fairer, more confident decisions
✓ Fits your existing interview workflow
✓ Human review stays in control
Ready to review deepfake risk with clarity?
See how Sherlock turns an authenticity concern into reviewable evidence.
VIDEO ANALYSIS FAQ
What can I upload?
Does this work for AI interviews and human-led interviews?
What does the API return?
Can reviewers jump to the evidence?
Does Sherlock make the hiring decision?
Upload a video through the API and see the evidence Sherlock returns.