daviidsusanto / smarterp

ERPNext App: Improvements for https://github.com/frappe/erpnext with data analysis and machine learning.

Smarterp

This repository is under development. The aim is to improve usability of the ERPNext (and Frappe) Open Source enterprise resource planning web platform (https://github.com/frappe/erpnext). I started working on this during the corona crisis and am looking forward to add additional functionality as by request from others or by my own terms. The main focus of additional features is based on dashboarding, data analysis and machine learning algorithms.

The packages that are installed with this frappe app as per requirement are sklearn, pandas and numpy. If you want to install the app, try it out, have requests of what feature could come in hand or want to collaborate please feel free to contact me.

Content

Screenshots

Included:

• Assignment Assistant for the Issues (Module: Support of ERPNext)

Roadmap:

• Add option to make assignment automatically without user interaction
• Visualized Loadbalancing for the Auto Assignment based on the open ToDo's of people
• Adding NLP to the Auto Assignment
• Finding similar documents for certain doctypes ...

Assignment Assistant

The model, that predicts the percentages is trained with existing data in your ERPNext Installation. You should have at least ~ 100 issues in your erp. It is based on the doctypes:

• Todos
• Issues
• Timesheets
• Projects
• Customers
• Employees
• Contacts

If you need to exclude certain doctypes because you don't use them, feel free to contact me. We can include a selection of doctypes in the settings of the app.

All of the doctypes should contain content in your installation. The more the better. If your testing or development environment does not support live data, you can in the settings use the login information for a production environment. It will pull data from there in a one-way fashion. It will not delete, modify or otherwise compromise the production environment. Please also check with your data privacy. The data will be downloaded as tables onto the system where you install the app. It will be located under /sites/sitename/assigner You can always delete this folder and all data is removed.

Installation

Just as any other frappe app:

bench get-app [this_repo_link]

bench --site [site-name] install-app smarterp

Configuration

1. Go to your ERPNext website
2. Reload (if necessary)
3. Type in search field: Smarterp Settings
4. Enter the Administrator username of your ERPNext instance (use a production instance, it will pull data from there (one-way) )
5. Enter the Administrator password (use production instance)
6. Enter the base url of your site (without trailing slash / use production instance)
7. Save the document
8. Click "Train Assigner" (Takes about 10 minutes, Reload the page and the Last Trained field will be set after training has finished)
9. For better performance the Assigner is trained every day around 5 a.m. automatically

Contact

[email protected]

License

MIT