Litera Collaborate is a Web App with features like Drive in which folder and files could be created or uploaded og unlimited size, Emailing functionality, Saring of files and folders with other persons with read or edit rights, multi-tenancy architecture, Authentication using SAML, LDAP, JWT and DRF token. The Application has just undergone Internal release and currently under Alpha Test with two of the company's trusted customers.

The web server was developed using Django and Python which provides the API that is send to Binary in Boards which then run the specific program. The further need was to enable analytics on the server where all the boards provide their input.

To create an application which would analyse the trends in market and recommend the client purchase and sales. Solution: To analyse the trend in the market it was needed that a supervised learning algorithm need to be implemented. The data was gathered from different locations of client business and was analysed to find the patterns about the sales of various goods from different locations. As per the analysis the client was suggested about what goods need to be purchased in what quantity and what location could achieve better business by concentrating more on a particular set of goods in a specific area.

To create an application using supervised learning algorithm which could predict the maximum revenue for the client.

The solution requires that an application should be generated which uses supervised learning algorithm which can predict the time during which the purchase and subsequently sale of any item will generate the highest amount of revenue, with penalized regression time duration which provide better ranking if the duration between purchase and subsequent sale of the item is less.

To create an interface with bank to pay to the vendors

Solution: Payment was to be done from SBI account of the dealer. To process the payment bank needed a details in text file format by mail. I designed and developed the text file in desired format from the information of vendor bank account and dealers bank account stored in database. The frequency of file to be generated was once a day after 6 PM.

To create an application which could predict the risk portfolio of customer for extending the credit line

The solution requires analysis of historical data provided by customers and their performance in repaying the credit line forwarded to them. The prediction of future risk associated with any customer is calculated on basis of a supervised learning algorithm which could predict the behaviour of customer by analysing the model derived from training the sample dataset with different weights associated with different parameters and upon the customer scoring more than a certain threshold.