🚀 About Me

🧑‍💼 Currently working as a Data Scientist at the 4th largest communications company in my country.

🎓 Background in law and communication consulting for various companies.

💡 Skills & Expertise

💻 Proficient in Python, SQL, and Power BI.
📊 Experienced in data cleaning, engineering, and visualization.
🤖 Knowledgeable in agile methodologies, workflow processes, and model evaluation.
🚀 Familiar with ETL processes, machine learning, and MLops.

🌐 Connect with Me

📧 Email: [email protected]
🔗 LinkedIn: https://www.linkedin.com/in/hasalazars/

This sentiment analysis project was carried out by performing tasks inherent to data engineering, data analysis, and machine learning engineering. Various data sources such as Google Maps, Yelp, and hotels.com were utilized, followed by data cleaning and loading into a data warehouse. Additionally, an incremental query and loading process were implemented, complete with the necessary data functions and orchestrator.

The project was implemented on the Google Cloud Platform. Python libraries such as Streamlit, Seaborn, and Matplotlib were used for visualizations, and a Power BI dashboard was developed. Furthermore, several natural language processing models were implemented to extract sentiment from reviews. Lastly, the OpenAI API (chat-GPT) was integrated to run the model with higher accuracy.

This sentiment analysis project was conducted in collaboration with four other colleagues:

• Douglas Sánchez
• Yaneth Ramírez
• Cristhian Castro
• Rodrigo Moreira

This repository showcases a comprehensive full stack Data Scientist process, encompassing various stages from data engineering to machine learning and deployment.

📚 Contents

1. Data Engineering: Collecting and transforming the data

• ETL (Extraction, Transformation and Loading)
  • Data Types
  • Missing Values
  • Unnesting data from columns in the dataset
  • Cleaning Data for future use
  • Loading Data in a structured format
• Developing API: 6 API functions with FastAPI

2. Machine Learning: Analyze and Model Training

• EDA (Exploratory Data Analysis)

  • Knowing the Data Set
  • Missing Values
  • Comparative Analysis with Graphics
  • Outliers
  • Word Clouds

• Movies Recomendation System: Model implemented with an API function

  • recomendacion(titulo): It enters the name of a movie and recommends similar ones in a list of 5 values.

3. Deployment: The model is deployed in render.com
   • Try the system Here - Movies Recomendation System

This project encompasses an Exploratory Data Analysis (EDA) conducted for a telecommunications service provider. The primary goal of this analysis was to gain insights into the behavior of the telecommunications sector at the national level. The aim was to provide valuable information to enhance service quality, identify growth opportunities, and propose tailored solutions for potential clients.

The project includes the following key components:

EDA Report in Jupyter Notebook: A comprehensive EDA report has been generated in a Jupyter Notebook. This report delves into the dataset to extract meaningful patterns, trends, and correlations. By utilizing various data visualization techniques, the analysis sheds light on the industry's dynamics, customer preferences, and potential areas for improvement.

Power BI Dashboard Development: Alongside the EDA report, a dynamic and interactive Power BI dashboard has been created. This dashboard visually presents the findings of the analysis, providing a user-friendly interface to explore data insights and key performance indicators. It empowers stakeholders to make informed decisions and respond proactively to market demands.

By combining the EDA report and Power BI dashboard, this project equips the telecommunications company with valuable information to enhance their services, capitalize on growth opportunities, and cater to the unique needs of potential customers. The insights gained from the analysis pave the way for data-driven strategies and improved customer experiences, driving the company's success in a competitive market landscape.

This project is a process of data cleaning, insights discovery, and visualization made from two datasets with user information. The datasets are from users who signed up for a talent recruiter platform.