FOLIO — V2.0/PUNE, INDIA

OBSERVER

ROLEProduct Design + Full Stack Development
STACKGo, React, Wails, Python, ONNX Runtime, SQLite, Tailwind CSS, Computer Vision, Machine Learning, Concurrency, Desktop App, Wails
YEAR2026

A local-first AI-powered desktop application that automatically understands, organizes, and searches image libraries without relying on cloud services.

[ drop a render → /public/ascii ]
[FIG.1]P01 — HERO FRAME

Observer scans image collections and generates AI-powered semantic tags locally using an ONNX image classification model. Users can instantly search thousands of images using natural keywords without manually organizing folders.

The application is built with Go and Wails for native desktop performance while Python handles machine learning inference. A concurrent worker system processes folder scanning, thumbnail generation, metadata indexing and AI tagging simultaneously to keep the interface responsive even for very large libraries.

Privacy was the primary design goal. Every image stays on the user's machine—no uploads, cloud APIs or external services. Observer demonstrates desktop application architecture, concurrent programming and practical machine learning deployment in a real-world product.

[ drop a render → /public/ascii ]
[FIG.2]DETAIL — 01
[ drop a render → /public/ascii ]
[FIG.3]DETAIL — 02