Image Intelligence Studio
One studio for image forensics, OCR, background removal and metadata scrubbing.
- Architect and full-stack developer
- 2026

What it is
A web studio for computer-vision work — steganography detection, authenticity analysis, OCR, background removal, restoration and metadata scrubbing — with per-user history, batch mode and chained operations.
The problem
Image operations range from instant to minutes long, need Python's computer-vision libraries, and handle files users may not want kept.
Architecture
Requests flow left to right. Hover a component for its role.
- React studio client, Vite, TypeScript
- Nginx edge
- Spring Boot service, identity, jobs, public API
- PostgreSQL and Redis data
- RabbitMQ queue, long jobs
- FastAPI CV service worker, internal only
- OpenCV, Pillow, Tesseract ai
Key decisions
Each runtime owns what it is good at
Spring Boot owns identity, authorisation, job orchestration and the public API. The Python service owns everything touching OpenCV, Pillow, NumPy or ML models and is never exposed to the internet.
Fast path and slow path
Quick operations are synchronous REST calls; long ones become RabbitMQ jobs with progress pushed over WebSockets.
Files expire
Uploads are checked by magic bytes and deleted after three days by a cleanup timer.
Highlights
- Built in eight milestones, from schema to systemd deployment.
- Batch mode and operation chaining on top of a single operation registry.
- OTP email sign-up and rotating refresh tokens.