https://github.com/obscurely/rekosearch-public
AI-powered search engine that understands the content of your files
https://github.com/obscurely/rekosearch-public
ai aws flask kubernetes python react rust search terraform
Last synced: 3 months ago
JSON representation
AI-powered search engine that understands the content of your files
- Host: GitHub
- URL: https://github.com/obscurely/rekosearch-public
- Owner: Obscurely
- License: other
- Created: 2025-08-16T09:51:43.000Z (11 months ago)
- Default Branch: master
- Last Pushed: 2025-09-01T11:57:38.000Z (10 months ago)
- Last Synced: 2025-09-01T13:38:23.453Z (10 months ago)
- Topics: ai, aws, flask, kubernetes, python, react, rust, search, terraform
- Homepage: https://rekosearch.com
- Size: 4.56 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- Funding: .github/FUNDING.yml
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
- Security: SECURITY.md
- Support: docs/SUPPORTED.md
Awesome Lists containing this project
README
RekoSearch is an AI-powered file search engine for images, videos, documents and audio that understands the content of your files. It enables simultaneous semantic search across them.
Explore the docs ยป
View Demo
ยท
Report Bug
ยท
Request Feature
Table of Contents
-
๐ชฝ About The Project
- ๐๏ธ Architecture Diagrams
- ๐ Contributing
- โ FAQ
- ๐ฐ Security
- ๐ชช License
- ๐ง Contact
## ๐ชฝ About The Project
RekoSearch is an AI-powered file search engine for images, videos, documents and audio that understands the content of your files. It enables simultaneous semantic search across:
- **Images & Videos**: Find objects, scenes, activities, landmarks, facial attributes and text.
- **Documents**: Search through the actual content (text).
- **Audio**: Search through what's being said.
You can plainly search for what you want, or use advanced search syntax like the example below:
```
label:dog AND face:smiling NOT text:warning
```
(This will find images and videos with a dog and a person smiling, but no text spelling "warning".)
Alternatively, you can also download just the results files.
If you've got any questions please give the [FAQ](#faq) a check.
### ๐ก Features
- [x] AI-Powered Content Search
- [x] Multi-Format Support
- [x] Advanced Search Syntax
- [x] Object and Scene Detection
- [x] Facial Analysis
- [x] Text Recognition
- [x] Audio Transcription
- [x] Batch Processing
- [x] Search Result Downloads
- [x] Secure Data Handling
- [x] Job Management
- [x] Transparent Pricing
### ๐ Demo
Check out the [demo](https://rekosearch.com/demo) to see how it works in action.
### ๐ Built with
- [AWS](https://aws.amazon.com/)
- [Kubernetes](https://kubernetes.io/)
- [Rust](https://www.rust-lang.org/)
- [Python](https://www.python.org/)
- [Flask](https://flask.palletsprojects.com/)
- [React](https://reactjs.org/)
- [MUI](https://mui.com/)
- [Terraform](https://www.terraform.io/)
## ๐๏ธ Architecture Diagrams
**IMPORTANT NOTE**:
- What is below referred to as the **FRONTEND** includes **ONLY** the **DASHBOARD**.
- The **HOMEPAGE** is a statically exported (**prerendered**) **Next.js website hosted on S3** behind CloudFront. (using the same technologies as the Dashboard: React, MUI, etc.)
- What this means is that the **homepage is very inexpensive to run** and infinitely scalable, at a cost that is easy to bear, while the **dashboard**, which costs more to run, **will be paid for by the users**.
### High-Level Overview

A comprehensive overview of RekoSearch's architecture, showing how authentication, job processing, data upload, and result retrieval work together. [View detailed documentation](docs/OVERVIEW.md)
### New Job Flow

Step-by-step process of creating a new file processing job, from user request through authentication, credit checking, S3 upload generation, and job queuing. [View detailed documentation](docs/Application/NEW_JOB.md)
### Server Processing

Detailed workflow of how jobs are processed on the server side, including file processing, AI analysis, error handling, and status updates. [View detailed documentation](docs/Application/SERVER.md)
### User API

Architecture of user-related API endpoints including profile management, user updates, terms acceptance, and account deletion processes. [View detailed documentation](docs/Application/USER_API.md)
### Jobs API

Structure of job-related API endpoints for listing jobs, retrieving job details, accessing file URLs, and deleting jobs. [View detailed documentation](docs/Application/JOBS_API.md)
### K3s Cluster

Infrastructure diagram showing the K3s Kubernetes cluster deployment on AWS with load balancers, auto-scaling, and CloudFront integration. [View detailed documentation](docs/Application/K3S.md)
NOTE: Currently using Digital Ocean K8s cluster to slowly integrate the K3s cluster in. Making sure the launch is smooth and without issues. Might actually stick with DO K8s cluster for some time as at the current scale it's cheaper and easier to manage. K3s cluster will get cheaper, but at a higher scale.
## ๐ Contributing
While the actual source code is private if you have any suggestions feel free to open an [Issue](https://github.com/Obscurely/RekoSearch-Public/issues) about them.
Please read [CONTRIBUTING.md](CONTRIBUTING.md) and
[CODE_OF_CONDUCT.md](CODE_OF_CONDUCT.md).
## โFAQ
See [FAQ.md](./docs/FAQ.md) for frequently asked questions.
## ๐ฐ Security
If you happen to find any issues please carefully read
[SECURITY.md](SECURITY.md) and strictly follow everything stated.
## ๐ชช License
Is licensed under [Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)
## ๐ง Contact
Either post an issue in the
[Issues Tab](https://github.com/Obscurely/RekoSearch-Public/issues) or contact
me at this email address if you have more to say:
[contact@rekosearch.com](mailto:contact@rekosearch.com)
[](https://ko-fi.com/K3K3H29LV)