Krishnendu Bhowmick

Experience

• 

  University of Calcutta, Kolkata

  Jun 2017 - Aug 2017

  Biochemical Estimation of Micro Algae With Special Reference To Biodiesel Production
• 

  Indian Institute of Technology, Bombay

  Aug 2019 - Apr 2020

  Research Intern

  : My project was "Detection and Segmentation in Multi-organ nucleus H&E stained Histopathology Images using Fully Convolutional Neural Network".: Multi-Organ Nucleus Segmentation was used as a dataset.: For segmentation Unet, SegNet model architecture was utilized.: Python with the PyTorch framework was used for my research work.

• 

  MedTourEasy

  Sept 2020 - Oct 2020

  Data Analyst
• 

  Innerwork Solutions Pvt Ltd

  Oct 2020 - Mar 2021

  Business Analyst
• 

  Planet NEXTgen Technologies

  Jun 2021 - Nov 2024

  1. Developed 200+ Stored Procedures and optimized SQL queries, which reduced query execution time by 15% and improved report generation efficiency by 20%.2. Processed and maintained 6M+ rows of data monthly, optimizing ETL processes in Qlik Sense to improve load time by 10%.3. Designed and developed 20+ dashboards in Qlik Sense and Power BI, incorporating valuable Key Performance Indicators (KPIs), enhancing data-driven insights and directly impacting business operations.4. Automated data cleaning, validation, transformation, and repetitive processes using Python and SQL, eliminating manual tasks and saved over 90 hours.5. Worked closely with stakeholders & cross-functional teams to understand business requirements. 6. Led and managed a team of 5 analysts, monitored daily and weekly progress, and provided reports to management. Show less

  • Analytics Engineer

    Dec 2021 - Nov 2024

  • Analytics Engineer

    Jun 2021 - Dec 2021
• 

  Indian Institute of Technology, Bombay

  Nov 2024 - now

  Research Intern

  Project: Data-Driven Design of Novel Drug Molecules Analysed and transformed molecular data into SMILES notation and extracted features (numerical vectors) using ChemBERTa embeddings for further analysis. Applied contrastive learning to refine the embedding space by clustering molecules with similar IC50 values. Currently, implementing a Gaussian Mixture Model on the embedding space to model the distribution of existing molecules. The sampled embeddings will be passed to a decoder to generate new SMILES (novel molecules). Following this, docking and IC50 prediction will be performed on the newly generated candidate molecules. Show less