Abdul Mussavir

Data Analysis & Data Science | Turning Data into Scalable AI Solutions | RAG, Embeddings & Automation | Python, SQL, Power BI | Building Agentic Tools for Real-World Impact. Detail-focused Data Analyst with knowledge in data warehousing, process validation and business needs analysis. Proven to understand customer requirements and translate into actionable project plans. Dedicated and hard-working with passion for Big Data.

• Used Microsoft Excel to develop inventory tracking spreadsheets.
• Made recommendations on business intelligence tools to cut processing time by 15%.
• Improved relationship with customer by developing a unique business intelligence solution to meet complex requirements.

Objective:
Build an AI agent that can parse, understand, and answer questions from financial reports using Retrieval-Augmented Generation (RAG) techniques.

Tech Stack:

• n8n (no-code workflow automation)
• OpenAI for text embeddings
• Pinecone for vector search and storage
• OneDrive as the data source

Key Features:

• File ingestion from cloud (OneDrive, scalable to S3/Google Drive)
• Text chunking and semantic embedding
• Indexed and stored in Pinecone by namespace
• Real-time query interface that responds to financial questions
• Modular structure, scalable across enterprise cloud platforms

Use Case:
Designed for finance teams, data analysts, and consultants who work with financial documents like 10-Ks or quarterly reports, and need instant answers from within lengthy documents.

Future Enhancements:

• UI dashboard for stakeholder interaction
• Multi-file ingestion with automated vectorization
• User-based access to analytics

🧠 A fully functional backend-ready RAG pipeline—designed to save hours of manual work.

🔗 Want to try it? [Demo available on request]

Amazon Review Analytics: Sentiment, Length & Helpfulness (highlights features)

• Amazon Product Review Sentiment Analysis
  Analyzed Amazon product reviews to uncover customer sentiment patterns, review characteristics, and correlations between helpfulness votes, review length, and sentiment scores.
  
  Key Highlights:
  
  Processed and cleaned raw Amazon review data for analysis.
  
  Applied VADER Sentiment Analysis to classify reviews as positive, negative, or neutral.
  
  Engineered features such as review length and helpfulness ratio for deeper insights.
  
  Created visualizations including:
  
  Correlation matrix between review features.
  
  Review length distribution and sentiment comparison.
  
  Sentiment label distributions (text-based vs. proxy labels).
  
  VADER compound score distribution.
  
  Tools used: Python, Pandas, NumPy, Matplotlib, Seaborn, NLTK (VADER).
  
  GitHub Repository: https://github.com/mussaussie/Amazon_sentiment_Analysis
• Skills: Jupyter · Python (Programming Language) · Matplotlib · NLTK (VADER Sentiment Analysis) · Natural Language Processing (NLP) · Feature Engineering · Data Visualization Exploratory Data Analysis (EDA) · Python Pandas NumPy Matplotlib Seaborn NLTK (VADER Sentiment Analysis) · Seaborn · NLTK (VADER Sentiment Analysis) · Feature Engineering · Data Visualization Exploratory Data Analysis (EDA)

Store Trial Analysis with Python: Measuring Sales & Customer Lift“In this project I performed an end-to-end A/B analysis of a three-store pilot (Stores 77, 86, 88) against matched control stores. Using Python (pandas, NumPy, Matplotlib, SciPy),
• Aggregated monthly sales, unique customers, and transactions/customer
• Selected controls via pre-trial Pearson correlation + magnitude distance
• Conducted t-tests and plotted confidence intervals to identify significant lift
• Decomposed incremental sales into new-customer vs. basket-size drivers
• Compiled interactive Jupyter reports and visual dashboards
Key takeaways: Store 77 saw sustained +60 % sales lift driven 70 % by new customers; insights guided a full rollout strategy.”

link :- https://github.com/mussaussie/store-trial-analysis

Skills: Python (Programming Language) · Data Analysis · Statistical Data Analysis · Data Visualization · Data ModelingSkills: Python (Programming Language) · Data Analysis · Statistical Data Analysis · Data Visualization · Data Modeling

Web Scrapping, Regex and PandasWeb Scrapping, Regex and Pandas

• In this project, I utilized web scraping techniques to extract and analyze the text of Martin Luther King Jr.'s famous speech from an online source. By leveraging Python libraries such as BeautifulSoup and requests, I was able to retrieve the speech content from the webpage. I then processed the text by cleaning and normalizing it, removing punctuation, and converting it to lowercase. Finally, I used the pandas library to count the frequency of each word in the speech and saved the results to a CSV file. This project demonstrates my skills in web scraping, text processing, and data analysis using Python.

https://github.com/mussaussie/webscrapping/tree/main

English, Hindi, Urdu

Cricket