Avik Banerjee

PhD Topic: Optimizing the gas consumption of blockchain smart contracts

Explored methods to classify PDF documents based on their content, using document embedding techniques and neural networks.

- Worked in NodeJS and ReactJS on a portal for ideation and collaboration on robotics projects. Worked on both back-end and front-end components. - Worked on an Apache FLink pipeline aimed at including a machine learning model to process streaming data, using Deeplearning4J - Worked on a web-app, using ReactJS frontend and NodeJS backend to enable a user to create and run a neural network through the app GUI. The backend for the machine learning model and its training were implemented with TensorflowJS.

I worked on broker back office software using Java, Spring, JSPX, ActiveMQ and related technologies and developed message conversion capabilities for an enterprise integration gateway for communication between the broker back-office software and external systems.

Topic: Optimizing the gas consumption of blockchain smart contracts

Detecting the ambiguity in the first search query raised by a user to a conversational search system, using Transformer based language representation and Graph Convolutional Networks.

- CGPA 1.5 - Specialization in Autonomous Systems, covering courses such as Computer Vision, Correspondence Problems in Computer Vision, Machine Learning, Speech Recognition and Synthesis, and Deep Learning. - Final thesis: Detecting Ambiguity in Conversational Systems, using Transformer-based language representation and Graph Convolutional Networks.

- CGPA: 9.28/10 - Covered courses including Computer Organization and Architecture, Database Management Systems, Automata Theory, Operating Systems, Computer Networks and Image Processing. - Final thesis: Automatic Detection of Diabetic Retinopathy using image processing and machine learning tools on retinal fundus images.