Hey! I'm Faraaz Arsath

Master Data Science & IITM Advanced Programming Professional

Visit my linkedin page here.

I'm a passionate Data Scientist and experienced business manager who thrives on driving innovation and delivering impactful solutions through technology. I specialize in bridging the gap between business needs and technical execution, ensuring seamless integration of processes and systems. My focus is on creating user-centric solutions that enhance productivity, foster collaboration, and streamline workflows.

Technology Stack

I analyze data, build pipelines, develop machine learning models, create dashboards, manage projects, and automate workflows to deliver impactful solutions efficiently.

  • SQL, Python
  • Power BI, Advanced Excel
  • MySQL, PostgreSQL, Snowflake
  • Machine Learning
  • Natural Language Processing
  • Always up for learning something new!

My Projects

My work includes building predictive models, optimizing workflows, and creating interactive dashboards for analyzing large datasets. Additionally, I have experience with tools like Snowflake for data handling, ensuring efficient data storage and retrieval for complex analyses.

Building an Interactive Interview Question Chatbot App

As a Junior Data Scientist worked in this Omdena project for building a chatbot which is designed to enhance and streamline interview for candidates and leveraged LLM, prompt engineering and Retrieval Augmented Generation (RAG). Conducted data preprocessing, model fine-tuning, and validation to ensure optimal performance.

Customer Segmentation - E commerce Purchase

Developed a customer segmentation model using Natural Language Processing (NLP) techniques and K means clustering on an E-commerce database, resulting in grouped customer segments based on purchases and spending patterns, while also identifying top sales countries and product categories through exploratory data analysis (EDA).

Ensemble Techniques - Breast Cancer Detection

Implemented ensemble models for Breast Cancer Diagnosis, achieving F1 score accuracies of 94%, 90%, and 92%. Demonstrated expertise in statistical analysis, emphasizing Type II error. Utilized confusion matrix, ROC, and DET curves for performance evaluation. The Voting Classifier exhibited superior predictive power, outperforming other ensemble techniques.

Contact Me

Email: faraazarsath@gmail.com