Our client was delivering hardware to the customer. Sometimes thesehardwares is not working so our client will send Field Service Employee to fix it. While going there they may meet with an accident, or the location theyare going to may have some harmful condition that will result in some incident.

We were having some already collected data of the incident. Based on latitude and logitude we caterozie the incident based on reasons like Pet Alert, Bad Weather Alert, Slippery Alert conditions, etc.

Our client will schedule the employee assignment for next day. Our code scheduler will run during night time and check all the assignments, based on the location of assignments it will check what all incident has already occurred on that area. Find the category of it and save it as an alert for next day schedule. Whenever FSR will log in to system he/she will see the alerts for wherever they are going.

Our client wanted to automate the process of Incident Management whenever his employee is meeting with an accident.

A lot of tools were available in the market but all are quite manual like if your four-body part got hurt then for each one you will select the dropdown individually.

In our case whenever an incident is created, a text box will appear on the screen. You type the incident description in normal English language (need not be any particular format), the NLP will run on the text and tag the word with the BODY_PART, REASON, LOCATION, DATE tag wherever its matching.

We used StanfordCorNLP for tagging, ColumnDataClassfifier as the classifier. This classifier aspect some input files which will have the tagging information. We were having old data recorded by our client, we categorized the data with proper tags and kept it as the input file. The input file update was an iterative process, if we found any new word which we are not able to tag then it was added to that input file.

If weask some educated person to set up a password for your banking system you would set a strong password. But if weask this to an illiterate person it would be difficult for him/her, as they may not know how to use the keyboard or the importance of strong password.

I developed a passwordless banking system POC where the communication between client and server was happening based on public and private key encryption algorithm. Client will download the app from our server and share his public and secret key with symmetric key encryption. Once shared secret key is deleted, thus making clinet side secure.

Whenever we do load testing a hue log file will be generated, its difficult every time to parse this file manually and checkthroughput-response time andbottleneck resource if any.

I developed a tool where you can give the location of the file generated during load testing and mention the log format corresponding to it. The tool will automatically create an excel file as output, which will have the throughput and response time graph. Also it will metion if any resource was bottlenck during load testing.