This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend
What I Built
Have you ever remembered having something but completely forgotten where you kept it?
I built FindIt AI, a privacy-focused visual memory assistant designed to help with exactly that problem.
The idea is simple: take a photo of your desk, room, cupboard, backpack, or workspace, and FindIt AI uses local AI to detect objects and remember where they were seen.
Later, you can ask:
Where is my laptop?
FindIt AI searches its visual memories and returns the object's last seen location, detection time, confidence, and the evidence image showing where the object was detected.
It is intentionally described as a last seen location, because an object may have been moved after the photo was taken.
How it works
- Scan a space — Upload a photo of a room, desk, or workspace.
- Detect objects — Local YOLO object detection identifies objects in the image.
- Store visual memory — The detected objects, locations, timestamps, confidence scores, and bounding boxes are stored locally.
-
Ask a question — Search using a natural-language query such as
Where is my laptop? - Find the last seen location — FindIt AI returns the most recent matching memory with visual evidence.
The goal was to build something small but genuinely useful for everyday life while keeping personal photos local.
Demo
Here is the result after scanning a space and asking FindIt AI where the laptop was last seen.
Demo Flow
Photo → Local AI Detection → Visual Memory → Search → Last Seen Location
For example:
Where is my laptop?
→ Laptop
→ Last seen at Bedroom
→ 95% confidence
→ Evidence image with the detected laptop highlighted
Code
The complete project is open source:
GitHub: https://github.com/sohamsakhare76-ai/findit-ai
The repository contains the Flask application, object detection pipeline, SQLite database, search logic, frontend, tests, documentation, and security/contribution files.
How I Built It
FindIt AI is built around open-source AI running locally.
Tech Stack
- Python
- Flask
- Ultralytics YOLO
- SQLite
- Pillow / OpenCV
- HTML, CSS and JavaScript
Architecture
text
User Photo
↓
Local YOLO Object Detection
↓
Detected Objects + Bounding Boxes
↓
SQLite Visual Memory
↓
Natural-Language Search
↓
Last Seen Location + Evidence Image

Top comments (0)