Case Study — Computer Vision & OCR
ERP Screen Change Detector
An AI-powered computer vision system that automatically detects meaningful screen changes in ERP demo recordings, captures key screenshots, extracts OCR text, and generates structured JSON, CSV, and HTML outputs for business process documentation and SOP automation.
The Challenge
Documenting ERP business processes meant scrubbing through demo recordings frame by frame — pausing, screenshotting, and transcribing every screen manually. The result was slow, error-prone documentation that was already stale the moment the system changed.
What We Built
- Perceptual change detection with SSIM, ORB feature matching, and perceptual hashing so only meaningful screen changes trigger capture
- Automatic key-frame extraction from ERP demo recordings using OpenCV
- OCR text extraction with EasyOCR to capture on-screen content for every captured frame
- Jinja2 templates generating structured JSON, CSV, and read-friendly HTML documentation outputs
- YAML-driven configuration for per-system monitoring rules and capture thresholds
PythonOpenCVEasyOCRSSIMORBPerceptual HashingJinja2YAMLComputer VisionOCR
Results
Documentation time
Hours → minutes per recording
Change detection
SSIM + ORB + perceptual hash
Output formats
JSON · CSV · HTML
More Work
Batch Disclosure — Price-Moving Catalyst Detection
An AI-powered system that scans ASX and SEC filings for forward-looking disclosures — contracts, M&A, guidance — that move micro-cap prices before the market catches on.
WikiGen — Autonomous, Bias-Aware Knowledge Synthesis
A multi-agent GraphRAG system that autonomously writes reliable, bias-aware Wikipedia-style articles with verifiable provenance on every claim.
Let's Work Together
Build Something Like This?
This system solved a real problem. We can build something similar for you.