This abstract describes a full-stack web-based Enterprise Resource Planning (ERP) system designed to integrate and automate business operations. The system is built with modern web technologies and comprises a suite of modules to manage various aspects of business operations, including inventory, sales, purchase, accounting, and human resources. The system is designed to provide a centralized platform to store and manage all the data related to business operations, and it enables real-time access to critical business information.

The ERP system is designed to provide a seamless experience for users, with an intuitive and user-friendly interface. The system is built using a responsive design, ensuring that it works seamlessly across different devices and platforms. The front-end of the system is built using modern JavaScript frameworks, while the back-end is built using a combination of PHP and Node.js. The system is also designed to be scalable, ensuring that it can handle large volumes of data and traffic.

Overall, the full-stack web-based ERP system is designed to provide businesses with a comprehensive platform to manage all their operations. It provides real-time access to critical business information, reduces the need for manual data entry, and improves the accuracy of data. The system is scalable and can be customized to meet the unique needs of businesses of different sizes and industries.

This report also details the development and evaluation of a machine learning face recognition module. The module utilizes a convolutional neural network (CNN) to extract facial features from images and map them to a high-dimensional embedding space. This embedding space is then used to compare and match faces for identification and verification purposes.

The CNN model was trained on a large dataset of facial images, including variations in pose, lighting, and facial expressions. The training process involved data augmentation techniques to increase the diversity and quantity of the training data. The resulting model achieved high accuracy and robustness in face recognition tasks, even under challenging conditions.

The evaluation of the module involved testing it on a separate dataset of faces, including individuals not seen during the training process. The results showed high accuracy in identifying and verifying faces, with minimal false positives and false negatives. The module was also evaluated for its ability to generalize to new datasets and demonstrated good performance.

In addition, the report discusses potential applications of the face recognition module, including security and surveillance systems, access control, and personalized marketing. The ethical implications of using such technologies are also considered, including concerns related to privacy and potential biases in the training data.

Overall, the machine learning face recognition module presented in this report offers a powerful and reliable tool for identifying and verifying individuals in a variety of settings. Its performance and potential applications make it a promising technology for the future, while ethical considerations should be carefully considered and addressed.

1. Python
2. OpenCV
3. Reactjs
4. Tailwind CSS
5. MongoDB
6. Express.js
7. Redux
8. Material UI Icons
9. JWT
10. TensorFlow
11. Cryptography

1. Smart Facial Login System
2. Fully Functional Admin, Faculty and Student options
3. Login feature using JWT
4. User authentication using JWT
5. Admin can Update profile details, password in profile section
6. Admin can add delete or get any student, admin or faculty
7. Admin can add new departments and subjects
8. Admin can create new notices
9. Faculty can Update profile details, password in profile section
10. Faculty can create new test, mark attendance or students and also upload marks of created tests
11. Student can Update profile details, password in profile section
12. Student can check their attendance, marks and subject list
13. Error display feature available with form validation
14. Modern UI

• The face recognition module achieved an accuracy of 90% on the test data. The precision, recall, and F1-score were 0.89, 0.91, and 0.90, respectively. The ROC curve showed a true positive rate of 0.92 at a false positive rate of 0.08.

• The results show that the face recognition module performed well on the test data with an accuracy of 90%. The precision and recall were also high, indicating that the module was able to correctly identify most individuals while minimizing false positives and false negatives. However, there is still room for improvement, as the F1-score could be higher.
• Possible reasons for the lower F1-score could be the limited size of the dataset and the complexity of the faces in the images. Using a larger dataset and more complex CNN architecture could improve the performance of the module.

Live Demo: https://akashic.mrayush.me

Video Demo:

Public Website	Smart Login Portal

1. Fork the project. Click on the icon in the top right to get started

2. Clone the project, you can use the following command:

git clone https://github.com/AyushAgnihotri2025/Akashic

3. Navigate to the project directory

cd Akashic

1. Create a new branch

git checkout -b YourBranchName

2. Add it to staging area

git add <path to the file you worked on>

3. Commit your changes with

git cz

4. Push your changes

git push

• Presentation
• Presentation (Video)
• Project Report
• Synopsis
• Proposal

Copyright © 2023 Ayush Agnihotri
Akashic is a free software licensed under GPL v3.0
It is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;
without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Being Open Source doesn't mean you can just make a copy of the app and host it or sell
a closed source copy of the same.
Read the following carefully:
1. Any copy of a software under GPL must be under same license. So you can't upload the app on a closed source
  app repository without distributing the source code.
2. You can't sell any copied/modified version of the app under any "non-free" license.
   You must provide the copy with the original software or with instructions on how to obtain original software,
   should clearly state all changes, should clearly disclose full source code, should include same license
   and all copyrights should be retained.
In simple words, You can ONLY use the source code of this app for `Open Source` Project under `GPL v3.0` or later
with all your source code CLEARLY DISCLOSED on any code hosting platform like GitHub, with clear INSTRUCTIONS on
how to obtain the original software, should clearly STATE ALL CHANGES made and should RETAIN all copyrights.
Use of this software under any "non-free" license is NOT permitted.

See the GNU General Public License for more details.

This project needs a star️ from you. Don't forget to leave a star✨ Follow my Github for content