Venkatesh Chittineedi

The project was intended to identify and predict credit loan defaulters using customer information such as demographics, work patterns, residence status, and credit to debt ratio. Techniques used included segmentation and logistic regression. Responsibilities included extracting data from multiple sources, participating in continuous interaction with marketing and finance teams, converting raw data to processed data, achieving multi-dimensional segmentation analysis, and using logistic regression to obtain probabilities for non-defaulters and defaulters.

Worked on Computer Vision, NLP, and recommendation systems. Utilized PySpark and Delta lakes for distributed data transformations and handling large amounts of data. Conducted research on more accurate, fewer parameter models for faster deployment. Built NLP-based transformations, allowing the model to detect keywords from natural text input (e.g., 'Rename column A to B') to apply transformations over data stored in Delta lake. Implemented Auto ML to automate tasks like model selection, data processing for nulls/date-times/strings, and real-time predictions.

Designed and developed analysis systems to extract information from large-scale data. Developed machine learning projects on real-time data. Wrote Backend APIs in Flask and deployed models in the cloud. Stored data on MongoDB and Mysql databases. Built recommendation systems and an NLP Sentimental analyzer project.