The COVID 19 pandemic has taken a serious toll on all the people around the globe. This period was specifically difficult for the working class people. Many companies, taking advantage of the work from home concept and deeming it as a luxury, began throwing work and crunching the deadlines, thinking the productivity of employees is boosted while at home. This is not true. The pressure of deadlines, submissions and meetings have had a negative impact and in turn affected the overall output of many corporations. In India, people have only recently started taking mental health seriously. HR departments are particularly busy these days and are trying to devise the best strategy and policies which will benefit the employees and the company. The inspiration for Mind Mark arose from the following question:
- How does a company decide what policies to devise?
- What if there was a tool which would allow the company to consider employee input? Yes, currently the companies do utilize survey forms and take periodic inputs. But all those questions are purely objective or MCQ based. Take an example: If the survey form has the question: Are you satisfied with your salary? Yes or No. Now, an employee having a salary of $100K and an employee having salary of $20K will have the same reply (No (let us assume)). But, the severity varies in both cases. This cannot be judged by simple yes or no questions. The mindset behind the respective replies is not made clear.
The above factors inspired me to develop an application which can provide a detailed analysis of responses.
What it does
The users are presented with a set of questions. Based on their responses to the questions, a Satisfaction Score is generated. It also tells the primary topics on which the replies of the users are based. Mind Mark utilizes Expert AI's Natural Language API to predict the overall sentiment of the responses submitted by employees. Based on the analysis the corporation can devise or alter their policies in favor of both the employees and the company, maximizing the overall output.
How I built it
The Web Application was built by utilizing the functionality Natural Language API. Primary coding was done on Python. The Frontend was developed using HTML and CSS. Flask was used to communicate data between the machine learning model and the web app.
- First the functionality of the API was setup.
- HTML structure of the pages was designed.
- CSS styles were applied.
- Flask was setup. Data input and output processes were fixed.
- Fixed minor bugs.
- Deployed the application on Heroku.
Challenges I ran into
The main challenge which I faced was to integrate 3 different pages with each other and to parse the data between them. It took an entire day to understand the working of Flask, since it was my first time working on this technology.
Accomplishments that I am proud of
This was my first application in which I managed all areas of development. For this project, I learned HTML, CSS and the Flask framework. It took 8 hour efforts per day for 15 days for this project to be successful.
What I learned
- HTML CSS
- Natural Language API and its functions
- Deployment on Heroku
What's next for Mind Mark: Employee Satisfaction Analyzer
Next step is to include a database where the response and satisfaction quotient of all employees is stored. From the database, most talked topics will be filtered and presented. Some visual infographic like pi chart will be generated and an overall average satisfaction score of all the employees will be generated for an in-depth knowledge.