Live Demo

Evidence Processing Pipeline

Upload your own video clip or PDF document and watch the pipeline process it in real time. SHA-256 hashing, metadata extraction, and analysis stages run directly in your browser. One file per session — no account required.

Checking server…

Evident provides forensic analysis and document-preparation support. It does not provide legal advice, and all findings, draft filings, and reports must be reviewed by the responsible attorney, filer, or organization before use in court or agency proceedings.

Try It — No Account Required

Or choose a sample file:

Drop evidence file here or click to browse

Max 10 MB · 1 file · Video or PDF · All processing happens locally

1

SHA-256 Hash

Web Crypto API

2

Metadata Extract

File API + video probe

3

OCR

EasyOCR + Tesseract

4

Transcription

Whisper large-v3

5

Speaker Diarization

pyannote.audio

6

Object Detection

YOLOv8

7

Violation Engine

Constitutional rules

8

LLM Report

Local LLM analysis (Ollama)

Full Platform

13 Stages. Complete Processing.

The full Evident pipeline processes evidence through 13 deterministic stages — from cryptographic fingerprinting through AI-powered analysis to court-ready report generation. Each stage is auditable and reproducible.

Full 13-Stage Pipeline

Each stage shows which subscription tier unlocks it. Free accounts can process files through stages 1–4.

1

SHA-256 Hash

Cryptographic fingerprint computed from original bytes. Establishes chain of custody at intake.

Demo
2

Metadata Extract

Device data, duration, resolution, EXIF fields, and container format analysis.

Demo
3

OCR

Text extraction from video frames and PDF documents using Tesseract and EasyOCR.

Starter
4

Transcription

Speech-to-text with timestamped segments using Whisper large-v3.

Starter
5

Audio Events

Acoustic event detection, silence-gap analysis, and loudness profiling.

Essentials
6

Speaker Diarization

Speaker segmentation and turn labeling using pyannote.audio.

Professional
7

Object Detection

Frame-by-frame object and person detection using YOLOv8.

Essentials
8

Pose Estimation

Body posture analysis and force-cue estimation for use-of-force review.

Professional
9

Multi-frame Tracking

Subject tracking with persistent IDs across video frames.

Professional
10

Violation Engine

Constitutional and procedural rule checks against event sequences.

Professional
11

Timeline Synthesis

Chronological event reconstruction with cross-referenced citations.

Professional
12

Burden Analysis

Evidence weighting and legal argument support assessment.

Professional
13

Report Generation

Structured court-ready report with provenance links and audit trail.

Professional
Start Processing Your Evidence View Plans

What Each Tier Includes

Demo Free No account required

Try the pipeline in your browser. Local processing only.

  • SHA-256 chain of custody
  • Metadata extraction
10 MB limit · 1 file per session
Try Demo Above
Starter $29/mo Per user, billed monthly

Full ingest pipeline with transcription and export.

  • Everything in Demo
  • Audio extraction (ffmpeg)
  • AI transcription (Whisper)
  • On-screen text OCR
  • PDF / DOCX export
25 MB limit · 3 files per case
Create Free Account
Essentials $79/mo Per user, billed monthly

Speaker analysis, event detection, and object recognition.

  • Everything in Starter
  • Speaker diarization (voice ID)
  • Audio event detection
  • Object & person detection
500 MB limit · 15 files per case
Choose Plan
Professional $199/mo Per user, billed monthly

Full forensic suite with violation analysis and court-ready output.

  • Everything in Essentials
  • Pose & force estimation
  • Multi-frame tracking
  • Violation & sequence engine
  • Burden-of-proof analysis
  • LLM constitutional report
  • REST API access
  • Court document formatter
2 GB limit · 25 files per case
Choose Plan

Enterprise — Unlimited files, custom integrations, and dedicated support. Contact Sales