Saibal Patra's repositories
AI-Powered-Video-Content-Explorer-and-Query-Resolver
The project is a Retrieval-Augmented Generation (RAG) application that leverages OpenAI’s Whisper to transcribe YouTube videos and answer queries about the content. The tech stack includes Langchain, OpenAI, HuggingFace, Pinecone, and Google GenAI to ensure efficient transcription, question-answering, and data management.
Building-RAG-Application
Building RAG Application using Python, Llama2, Pinecone, langchain
SD-video-to-HD-video-convertor
To develop a prototype of a video conversion process that can convert SD videos to HD videos using diffusion (image to image, Inpainting) type models, while preserving the context of the video and being efficient.
DoorDash
Doordash - Predictive model for accurately estimating total delivery duration|Linear Regression,LGBM|Personal Project Build a predictive model for DoorDash to accurately estimate the total delivery duration in seconds from order submission to actual delivery, enhancing consumer experience.
Sign-Up-Flow-Optimization-Analysis-with-SQL-and-PowerBI
The sign-up flow optimization analysis focuses on using MySQL queries and Tableau for data analysis to identify issues in the registration process of 365's website visitors. It covers data visualization, A/B testing, and offers actionable solutions to enhance user experience, ultimately improving website performance and conversion rates.
Solving-the-Churn-Prediction-Problem-Building-a-Robust-Model
A business manager of a consumer credit card portfolio is facing the problem of customer attrition. They want to analyze the data to find out the reason behind this and leverage the same to predict customers who are likely to drop off.
Stock_Market_Analysis
In this project, we 'The Phoenix' will create a model that will help us to predict the stock price, with colorful visualization.
Analyzing-Banking-Trends-Customer-Transactions-and-Regional-Impact
In the dynamic world of banking and finance, understanding customer behavior and transaction impacts on regions is vital for decision-making. Our goal is to identify patterns that could influence regional economies and financial systems.
Case-Studies
Case studies will be performed using different statistical methods.
CO2-Emissions-Analysis
Creating a detailed analysis of co2 emissions.
Cohort-Analysis
Cohort analysis is a statistical method used to track and compare the behaviors and performance of distinct groups of users (cohorts) over time, helping businesses gain insights into user retention, engagement, and the impact of changes or strategies over specific time periods.
Data-Engineering
Data Engineering Repo
Dating-App-Reviews
Leveraging a vast dataset comprising over 600k dating app reviews spanning 2017-2022, I conducted a comprehensive analysis to uncover trends in user behavior and app performance.
Generative-Adversial-Network-Tensorflow
Generative Adversial Network using Tensorflow. FIrst, Generative Adversial Network --> Conditional Adversial Network --> Pix2Pix GAN --> cycleGAN --> SRGAN --> Semi Supervised Learning with GAN
Insights-to-the-Revenue-Team
Insights to the Revenue Team in the Hospitality Domain
Market-Analysis-using-PSQL
This project focuses on conducting a comprehensive marketing analysis using PostgreSQL. The goal is to derive valuable insights from a dataset of marketing interactions and customer behavior. Below, you'll find an overview of the project, its objectives, and key SQL queries used for analysis.
Natural-Language-Processing
Natural Language Processing and Resources
PostgreSQL
Problems solved from various recources
Projects-on-AWS
I will be creating projects on ML and deploy them using AWS.
Shopping-Data-Analysis
Track Data Analysis : Descriptive Analytics, Market Basket Analysis // Track Data Science : Regression, Forecasting, Customer Churn Prediction
SKYHACK-United-Airlines
As an analyst, you are required to leverage data to help in identifying opportunity areas in United’s current Food &Beverage(F&B) service and make recommendations which can help in increasing F&B (Food &Beverage) service satisfaction rate by identifying pain points for our customers and challenges in our current inventory planning.
Trash-to-Treasure-Uncovering-Municipal-Waste-Trends
This project aims to develop a data-driven waste prediction model using a comprehensive dataset with over 4,000 data points across 20 regions and 102 provinces. By leveraging this rich dataset, our project seeks to enhance waste management practices and promote efficient resource allocation.
Working-with-API
In this repository, I will collect data using API and then Analyse it.