We have developed three divisions of software:

  1. Enterprise Resource Planning (ERP) + CRM for Textile garment manufacturing
  2. Deep Learning (DL) applications: In the area of Cognitive Robotic Process Automation: Image recognition a) Option Identifier b) Survey results summary.           In Video recognition DL is trained to recognize people entering a de-marcatized zone
  3. Machine Learning applications:

a) Natural language processing of

  • Child abuse complaint records
  • User reviews of restaurants
  • Law bill reviews

b) Predictive Analytics of

  • Consumer patterns
  • Loan defaulters
  • Machine service requirement

We have developed a Natural Language Processing engine. This has been successfully applied to understanding informal English (user reviews) and for formal precise analysis (Law bill reviews)

I) Enterprise Resource Planning Software + Customer Realtionship Management for Textile manufacturing company
We have developed a complete Enterprise Resource Planning software for Textile Manufacturing companies. This is a cloud based application, using ASP.Net, MVC, jQuery. You can use this ERP to efficiently manage all tasks in a textile manufacturing company, right from

  1. Forecasting
  2. Projection
  3. Enquiry
  4. Costing
  5. Time and Action
  6. Sample Job order
  7. Bulk Job order
  8. CAD
  9. Yarn
  10. Woven Fabric
  11. Knit
  12. Fabric
  13. Trims
  14. Process quotation
  15. Production
  16. Production reports
  17. General Purchase
  18. Dyeing vertical
  19. Knitting vertical
  20. M.I.S reports
  21. Human Resources
  22. Logistics

You can access “ikvalTextileERP” application using the below link.

http://textileerpdemo.us-east-2.elasticbeanstalk.com/Account/Login

Please drop me a mail at arun@ikval.com, describing yourself /  your company, so that I can provide you with the user name and password. You can use the application to satisfy yourself of the features before we discuss your purchase of it.

An user creation and authentication ensures authorized usage.

Master tables for over Forty items are available along with Generic masters.

  • You can see an usage document of our ERP through this link

https://drive.google.com/open?id=1YE1bAK5biIrgWU5Bnwi4dUGQZbaWoh3e

  • We are now enhancing this with

a) Customer profiles and Customer Interactions and

b) Predictive analytics for Demand, Price, Machine service with Machine Learning to create an IntelligentTextileERP.

 

I) Cognitive Robotic Process Automation: Image recognition: Option identifier

Using Deep Learning, we built an application that scans PDF / JPEG documents and identifies

  • Options chosen by user: Yes or No for Ten sections. This application handled the problems of a) Multi-label classification b) Pose problem. The usage of the application reduced human labor and reduced human error.

Fig: Application in English and then in Spanish with Handmade marks to choose options.

Application identified Tick marks, Cross marks, Circle and Shade marks by the user.

  • Presence of Signature

Figure: Presence of Signature in application is identified

  • Presence of Handwriting anywhere in the document. If there were any handwritten messages then human processing was required. The application correctly identified such cases. Here a combination of OpenCV functionality and Deep learning is used.

Figure: Handwriting anywhere in the application is identified

The Deep Learning application that we built used

i) Inception model, ii)Convolutional Neural Networks.

III) Cognitive Robotic Process Automation: Image recognition: Survey summary

A second Deep Learning product recognizes the choices marked by user in answering a survey. A spreadsheet is generated of the results of all survey respondents. The department staff only needs to see the spreadsheet to analyze results of the complete survey.

FFigure: Survey document; results identified by application provided as a spreadsheet

IV) Natural Language Processing of Health complaints

We have built a Natural Language processing application that is able to read records in English and process them for medical descriptions. Our application provides a risk value for the particular health case. If this risk value and the associated to-do actions does not match the values proposed by the Health-worker, then the Health-worker is requested to justify the risk estimate. This ease supervisor approval of the risk estimate and further actions for that case.

This application was developed using Python in TensorFlow along with OWL.

V) Natural Language Processing of Law Bill Narratives

Law bills have over two hundred line items.  Each of the narratives in a line item has to be reviewed and checked for:

  1. Map narrative to UTBMS code
  2. DIs-allowed activities
  3. Block billing

Our NLP engine is able to identify entries in PDF of a bill and analyze the narratives to identify above three requirements. This results in a saving for the corporate. This is also done efficiently and in a repeatable standardized manner.

Bill reviewFigure: Analysis of law bill narratives, with savings per Line Item, per Unit, and in Dollars shown as Pie Charts.