Voice Analysis by embedding voice markers for Homeopathy Treatment
Developed an algorithm for splitting up audio clips to small segments, extracts audio features like Zero
Crossing Rate, Energy, MFCC, Spectral Centroid, (additional 17 such features) using Python library
called Librosa.
The same could predict the value of arousal and valence for a new audio clip using Numpy, Scikit-Learn,
Pandas.
With the values of audio features and corresponding value of arousal and valence, trained Support Vector
Regression Model
Arousal and valence of the audio file is calculated and thus the corresponding mood is predicted using a
kNN classifier
Built a customer interaction chat based plugin using web hooks and slack API which can be added to a CRM website.
Integrated Personality Insight API of IBM’s cognitive AI Watson to analyse customer satisfaction and generate customer reports
nvolved Slack as a channel to communicate with customers, as well as an independent real-time communication platform, with material design UI framework, hosted on MEAN
Created a web-app for AMC, a linux based software using HTML5, CSS3 and AngularJS on frontend and built on Django
Generate question papers, scan answer sheets, create mark sheet on the server and the results are returned to the web-application from the Django backend, python scripts contain shell commands are fired which then
Coded and analysed the different algorithm for predicting onset of Diabetes using the Pima-Indian
Diabetes by using various algorithms like -Naïve Bay’s, Random Forest and Logistic Regression.
Devised the prediction models on Jupyter Notebook.