Worked on to enhance customer search experience related to search results, type ahead suggestions and knowledge panel using NLP, Learning to Rank.

•Built Image classification system using Nilearn, Nibabel modules for 3D MRI ( .nii ) processing and performed skull extraction using HD-BET•Extracted 14 haralick features based on GLCM matrix using mahotas and achieved an accuracy of 72% by combining Naive Bayes and RF•Improved the overall accuracy by 10% through MRI image segmentation using U-Net architecture in Keras to demarcate tumor boundaries

• Employed NLP techniques such as Word Embedding using Glove, NLTK and Textblob etc. to classify social media text posts• Implemented Maximum Known Matching(MKM) scoring method and NLP libraries to split non-camelcase hashtags• Assembled Hashtag Splitting algorithm with text classification, used for Sentimental Analysis for a product launched by company • Tagged images posted on Social Media using YOLO v2 trained on 80 class MS COCO dataset for content cross promotion• Associated with team involved in building an Insurance chatbot product, using semantic similarity calculated via TF*IDF algorithm• Ensembled algorithms and launched the chatbot on Telegram app by the name WNS DS Bot

 Member of the apex body responsible for leading a team of 20 core organising team members and 25 associate members Spearheading more than 15 events of 25th edition of COMPOSIT, largest Annual Metallurgical fest of India Revamped the corporate communication modules and deliverables, aimed at improving corporate associations and fostering the growth of the fest Single-handedly responsible for the formulation, planning and execution of the music, literary and creative genre of the fest

 Developed an Android application giving real-time health condition classification of the site worker. Used OpenCV for multi-scale template matching in order to automate the complete warning system. Constructed a remote MySQL database for real-time linking of the central system and with the mobile application. Worked with instantly measured WBGT, heart rate & RSPM to calculate heat-stress level & air quality index. Devised a multi-label SVM Classifier model with hyper-parameter optimization by Genetic Algorithm. Predicted health condition of the workers using heart-beat of the employee, WBGT, RSPM value of the working area, temperature and humidity.

Developed a Decision Support System whose User Interface was built completely using Kivy in Python.The System performed Descriptive Analysis and Predictive Analysis with Evaluation of Models based on Performance Parameters like Accuracy, Precision, Recall, F1 score and AUC value and ROC curve.Implemented Genetic Algorithm for model optimization and applied it to the industrial dataset. Addressed questions raised on Risk Assessment Prediction, Causal Identification and Incident Prediction arised from datasets.

 Responsible for the Organised meetings, Practice sessions and workshops for all the 5 events of Dramatics Cup, Social and Cultural General Championship for the session 2016 – 17 Out of 22 Halls, Nehru Hall registered the historic win of Dramatics Cup, Social and Cultural General Championship for the session 2016-17 with podium in all the 5 events (2 Gold, 2 Silver and 1 Bronze) Individually managed the financial budget of INR 2.5 lakhs required for the successful organization and equipments of the events in the hall

 Individually garnered funds amounting to INR 2.08 lakhs through corporate deals and alumni contributions Oversaw the accommodation of 350+ guests (15% YoY increase) comprising participants, alumni and corporate guests Co-headed a team of 17 associate members, expanding the outreach of COMPOSIT to 100+ colleges and published over 20 media articles Spearheaded the publicity campaign in Odissa leading to a 20% YoY increase in participation