Python Developer – AI Forensic Analysis

Position: Senior Python Developer

Experience:

Shift:

Location:

About the Company

human authorship

video, image, text, and code

Role Overview

proprietary Internal Forensic Dashboard

strictly NO-AI-CODE

No ChatGPT, CoPilot, Cursor, or LLMs may be used to generate code.

We audit codebases, and any synthetic structures result in termination.

real engineers

Key Responsibilities

1. Video Forensics (Highest Priority)

• Implement FFmpeg-based

  frame extraction

  and keyframe sampling
• Analyze

  temporal consistency

  , pixel noise, and lighting variance
• Build

  Deepfake Detection

  using Error Level Analysis (ELA)
• Perform frame-by-frame artifact scanning

2. Image Forensics

• Implement

  Error Level Analysis (ELA)

  using OpenCV
• Build

  FFT frequency-domain analysis

  to detect GAN/Diffusion patterns
• Extract

  Metadata, EXIF, C2PA

  tags

3. Text Forensics

• Integrate Hugging Face models (e.g., RoBERTa) for:

• Perplexity scoring

• Burstiness measurement

• Implement

  stylometric analysis

  (sentence variance, structure patterns)

4. Code Forensics

• Build Git history “

  velocity analysis

  ”
• Parse AST Structures to detect AI-generated code blocks
• Identify non-human coding patterns

5. UI + Dashboard Development

• Build a

  local browser-based UI

  using Streamlit or FastAPI
• Enable asset uploads for all four modalities
• Display visual outputs (heatmaps, FFT plots, graphs)

6. Forensic Report Generator (Legal-Grade PDF)

• Build dynamic reporting engine using Jinja2 + ReportLab/FPDF/WeasyPrint
• Embed forensic visuals programmatically
• Add auditor overrides & human-in-the-loop notes
• Toggle evidence sections on/off

Required Skills

Technical Must-Haves

• Strong Python engineering experience (5+ years)
• Deep understanding of:
• NumPy
• OpenCV
• FFmpeg
• Computer vision math
• Experience processing

  large video & image files locally

• Strong foundational math: luminance analysis, gradient detection, spectral analysis
• Experience with

  Streamlit, FastAPI

  or similar frameworks
• PDF generation experience using:
• ReportLab
• WeasyPrint
• FPDF
• Strong debugging & system-level reasoning

Mindset Requirements

• Must write

  clean, human-authored, verifiable code

• No AI-generated code allowed
• Must provide GitHub links
• Comfortable working independently

Nice to Have

• Background in security forensics
• Experience with C2PA
• Experience with feature extraction from multimedia assets
• Familiarity with deepfake datasets