Avishek Mukherjee

Avishek Mukherjee,I work as a Data Scientist at Fidelity National Financial/ Servicelink (a subsidiary of Fidelity National Financial). I work mainly on Computer Vision problems as in Object Detection, Instance Segmentation, Video Analytics etc. with various kinds of engineering over anchor boxes, region proposal mechanism or bbox regression mechanism. With strong experience in OpenCV. I prune and quantize CNN models developed in Tensorflow or Pytorch before deploying them on cloud or on FPGAs. I have keen interest in GANs and in making classifiers more robust. I also work with Language Data (packages include: REGEX, SPACY, NLTK for simple works like NER, Text Normalization, Data Cleaning, Text Classification, Text Generation, Text Summarization models, I fine-tune existing models to our needs and Image Caption some images . Other functionalities include extracting some of the texts using OCR (Teseract). I also build light-weight statistical models, do A/B testing on hypothesis, clustering, feature extraction etc.. I worked on the engineering side as well where I deployed models to Azure cloud or Azure ML workspaces, and have inevitably dealt with Azure Queues and some NoSQL databases, Docker and Kubernetes.Prior to this I worked under Prof. Dr. Prabir Bhattacharya & Prof Dr. Siwei Lyu (now at SUNY Buffalo) where I applied AI, Computer Vision & Image Processing/Segmentation for Detection of Pathologies from Radiological Images. Prior to this I have some Software Engineering experience at Tata Consultancy Services, Larsen & Toubro Infotech Ltd.

Mainly configured for Data Science and AI jobs

• Image Recognition using Densely Connected Covolutional Neural Networks
• Detection of fake news using LSTMs
• Routing Measurement Across various geographies
• Simulating Gaussian Noise corrupting PAM and QAM signals in MATLAB
• Natural Language Processing, estimating gender-related use of pronouns using Parts of Speech Tagger by Stanford CoreNLP and parsing the XML file generated
• Sentiment Classification of Corona Tweets
• Using Word2Vec to generate Vector Representation & visualizing it with TSNE, alternatively I can also use GloVe
• Gradient Boosting Classifier for Gender Prediction on Disaster Data "Titanic"
• k-Nearest Neighbor and GridSearch on Diabetes Data with Feature Correlation Heatmap using Seaborn
• Movie Recommendation System using Collaborative Filtering built with Scikit-Surprise

I have experience in AI and Machine Learning R&D at a Fintech Company at Pittsburgh PA USA and at SUNY at Albany at Albany NY USA, and also at some IT services companies like Tata Consultancy Services, Larsen & Toubro Infotech where I worked as a IT Services Software Engineer in the domains of Property & Casualty Insurance and developed the TCS BANCS Genesys Configurator, the highest selling Insurance product in Indian Insurance market for business lines like Auto, Marine, Freight Forwarders Liability, Marine Hull, Fire Floater, Multi location industrial and have worked with IMS Health, now part of IQVIA in implementing their backbone Data Warehouse

Apart from Computer Science , I am an amateur guitar-player and painter. I at times play some Chess. And read poems by Johann Wolfgang von Goethe.