Project Info
This project did not submit a demo video on Devpost.
💡
Inspiration
💡 Navigating through a digital jungle of files and documents is a common struggle for students and professionals alike. The frustration of sifting through folders filled with cryptic file names like "syllabus_fall23 (5).pdf" or "README.txt" is something we all can relate to. The existing file management tools, like Google Drive, although efficient, often fall short in providing us with insightful, semantic information about our files' contents. This sparked our inspiration for a solution that could transform the way we organize and access our digital resources. ⚙️
What it does
⚙️ We present Semantic Drive, a cloud-based file management platform, akin to Google Drive, but with a revolutionary twist. We've harnessed the power of machine learning technologies, seamlessly integrated with MindsDB, to provide summary data for a wide range of file types. Whether you're dealing with text documents or images, Semantic Drive doesn't just store your files; it comprehends them. AI Track In the AI Track, we've taken file management to a whole new level by infusing it with the intelligence of machine learning. This involves using MindDB's integration with ML models as well as Hugging Face to allow for more accurate semantic searching. 🏗️
How we built it
🏗️ Our frontend was built with React and our backend was made using Flask. 🟣 CockroachDB Serverless 🟣 Our foundation is built upon CockroachDB Serverless, paired with Python, where we securely store your files and their associated metadata. This secure database forms the backbone of our system, ensuring your data is organized and accessible. We've created a user-friendly Flask backend to streamline the process of entering file data and metadata. This allows for efficient data management and retrieval. 🔴 MindsDB 🔴 MindsDB plays a pivotal role in the Semantic Drive ecosystem. As files are uploaded, MindsDB springs into action, automatically summarizing each file's content. What sets us apart is that this isn't limited to text documents alone; we've engineered a custom handler for MindsDB that can comprehend image data and other file formats, providing detailed, semantic descriptions. The semantic descriptions generated by MindsDB are seamlessly integrated into our Flask backend, enabling you to search for files using these intelligent descriptions, all while leveraging the power of Hugging Face. 🚩
Challenges we ran into
Throughout our journey, we encountered challenges that pushed us beyond our comfort zones. Exploring new technologies, such as MindsDB and Intel Cloud, presented a learning curve. However, it was a rewarding challenge, as it allowed us to expand our horizons and deepen our understanding of cutting-edge tools and platforms. 🥇
Accomplishments we're proud of
We've transformed the way files are managed and accessed. Semantic Drive's ability to understand and provide semantic descriptions of files is a significant achievement, making file searching more intuitive and efficient. 📚
What we learned
Our journey taught us how to harness the power of new APIs and technologies, enriching our skills and broadening our capabilities. ⏳
What's next
Semantic Drive is the proof that AI can enhance everyday tasks, like file searching, making them more intuitive and efficient. Our vision for the future includes expanding our capabilities further, removing the constraints of exact word searches and extending our semantic understanding to even more file formats, revolutionizing the way we interact with our digital resources. Semantic Drive is just the beginning of a smarter, more intuitive digital world. We hope that file managers like Google Drive will soon integrate a similar service.
Sembox
A drive with intelligent semantic searching.
For example, when you search for dog, it will bring up pictures, audios, texts... that contain dog.
Built (mostly) during Calhacks 10.0 (2023)
Components
Look at each component for more information
Potential Improvements
- better models
- custom summary models (especially text)
- Currently - size constraints for my hardware
- Also can get faster inference times with gpu
- better search model
- Can be much better - currently, just cosine similarity
- custom summary models (especially text)
- better ui
- light theme
- better metadata display
- lots of clean up (e.g. Loading)
- User management
Analysis
View
Metric
- 41
- 13
- 8
- 3
Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.
Technology
- FlaskIn code
- JavaScriptIn code
- Next.jsIn code
- PythonIn code
- ReactIn code
- PyTorchClaimed
5 of 6 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.
AI coding agents
No AI coding agent signals were found in this repository.
Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.
Codebase size
Source size
71 KB
Source files
25
Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.
Repository
sachiniyer/semantic-drive
43 files · 267 KB · @ ba04cdb
Structure
Interface
13 files · 30%Screens, components and styles rendered to the user.
Application logic
6 files · 14%Domain rules, services and shared utilities.
Supporting
Layers are inferred from where files sit in the tree, not from reading the code. A project that names its directories unconventionally will read oddly here — open the file browser to check anything the diagram implies.
Languages
- Markdown51%
- JavaScript29%
- Python19%
- Shell1%
Share of indexed source by file size. Binary and vendored files are excluded.
Dependencies
semantic-drive-frontend/package.json
npm · 10- @emotion/cache
- @emotion/react
- @emotion/styled
- @mui/icons-material
- @mui/material
- eslint
- eslint-config-next
- next
- react
- react-dom
semantic-drive-backend/requirements.txt
pypi · 8- flask
- flask_cors
- pillow
- psycopg2-binary
- python-dotenv
- requests
- sentence_transformers
- transformers[torch]
Declared in the repository’s manifests at the indexed commit. A declared package is not proof it is used, and runtime dependencies are listed first.
This project’s features have not been analysed yet.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.