Avisheka Guha Niyogi

• Analyzed large datasets to identify trends and patterns in customer behaviors for different natcos using PySpark in Router and STB fault data.
• Worked on clustering algorithms like Gaussian Mixture models to understand fault reporting patterns and predict fault reoccurrence among various user groups by analyzing technical parameters of device and connectivity.
• Applied cohort analysis on fault data to understand user behavior to capture reoccurring complaints statistics over a given year.
• Discovered stories told by data to present usecases to stakeholders.

• Working on a recommender engine that analyses past purchases of banking products made by clients during their relationship with the Bank and a series of additional socio- demographic attributes that will allow you to further adjust the purchasing profile. Through this information, we will predict the new financial products that will be preferred by each client.
• Conduct exploratory data analysis and necessary feature engineering for all assignments.
• Time Series Analysis to better understand the future direction and risks of various currencies against dollar. Hedge funds, get help with this analysis that gives them a consistent edge in predicting currency movements. We used both Time Series Analysis and Linear Regression. Used Azure ML and Azure Devops pipeline to deploy onto AKS.
• Created a model as proof of concept to check if Credit Risk is there for loans and deploy onto Azure Kubernetes Services using Azure Devops and Azure ML.

• Customer-Marketing-segmentation: We created model to segment a portfolio of credit card users according to usage characteristics. Customer segmentation is then used by marketing department to tailor products and services specific to these customer segments using KNN and PCA.
• Created a model as proof of concept to check if a telephonic campaign is working to sell insurance schemes.

Worked extensively on creating CI/CD pipelines using Jenkins where automated the deployment process from pulling code from Git, storing the artifacts in Nexus and using Ansible to configure and deploy them over servers in test environments and in Kubernetes cluster in AWS.

Used Jenkins and Maven for periodic job builds and maintain them.

Data and log analysis of transactions, file settlements using Linux.