Skip to content
View Arfa-Ahsan's full-sized avatar

Block or report Arfa-Ahsan

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this userโ€™s behavior. Learn more about reporting abuse.

Report abuse
Arfa-Ahsan/README.md

Hi there ๐Ÿ‘‹ I'm Arfa Ahsan

  • ๐Ÿ‘ฉโ€๐Ÿ’ป A Computer Science and Information Technology graduate.
  • ๐Ÿค– Passionate about Artificial Intelligence and Data Science.
  • ๐Ÿ” Deeply curious about how data and intelligent systems can change the way we live and work.
  • ๐Ÿš€ Aspiring to excel in AI and one day build my own startup that brings AI solutions to real-world challenges.

๐ŸŒŸ About Me

  • ๐Ÿ’ป Skilled in Python, SQL, Excel, and Power BI
  • ๐Ÿค– Experience with Machine Learning, Deep Learning, RAG, AI Agents, and Data Analysis
  • ๐Ÿ“Š Interested in Data Visualization, Business Intelligence, and Generative AI
  • ๐ŸŒ Love working on projects that create social impact and meaningful change

๐Ÿ“Œ Featured Projects

๐Ÿ”น RadioCheck-F1-Assistant An AI-powered Formula 1 assistant built with LangGraphโ€™s multi-agent system using a tool-based agent pattern. It features five specialized agents that;

  • Retrieve live driver and constructor standings through automated web scraping (via GitHub Actions)
  • Estimate championship outcomes using Monte Carlo simulations
  • Provide real-time weather analysis
  • Deliver official YouTube race highlights
  • Retrieve data through google search engine

All insights are delivered through a natural conversational interface, giving users a seamless, interactive F1 experience.

๐Ÿ”น Content Creation Agents An AI-driven multi-agent content generation system built on CrewAI. It automates the process of creating engaging, SEO-optimized content through three collaborating agents:

  • Media Analyst โ€“ Researches the given topic and identifies trending discussions.
  • Content Writer โ€“ Creates platform-tailored, engaging content.
  • SEO Optimizer โ€“ Refines the content for search engine optimization and better reach.

๐Ÿ”น Email Optimizer Agent
An intelligent email-writing assistant built using LangGraph and the Reflectโ€“Evaluator design pattern. The user provides an email topic (the more detailed, the better)

  • The Reflect (Generator) agent drafts the initial version of the email.
  • The Evaluator agent analyzes it based on parameters like clarity, grammar, call-to-action strength, and tone.

Developed with Groq LangChain, LangGraph, and MoonshotAI, this system helps users craft clear, professional, and persuasive emails through AI-guided refinement.

๐Ÿ”น UrduWhiz
An AI-based Urdu learning and reading platform designed for kids. It processes scanned Urdu storybooks (PDFs) and allows users to ask questions about the content interactively. Built using a Hybrid RAG (Retrieval-Augmented Generation) approach that combines keyword retrieval and semantic search. It further employs rerankers to improve document relevance.

๐Ÿ”น PII Detection & Redaction (Audio)
Many organizations such as banks, call centers, and utility companies record customer calls for quality or verification. During these interactions, users often share sensitive personal information like CNIC numbers, names, or account details. If left unfiltered, these recordings can pose serious data privacy risks.We have built an intelligent system that detects these sensitive entities and redacts them from the audio recordings.

๐Ÿ”น AI Wildlife Project
An AI-powered wildlife monitoring system integrating computer vision and audio signal processing. It uses YOLOv8 and Haar Cascade for animal detection and tracking, while Librosa processes sound patterns to identify species activity. The system aids research, conservation, and security by automating wildlife observation and behavioral analysis.

๐Ÿ”น Customer Churn Prediction
Developed a machine learning model to predict and explain why customers leave a service. Key steps included:

  • Data cleaning and exploratory analysis
  • Feature engineering and scaling
  • Handling class imbalance using SMOTE
  • Model training with hyperparameter tuning
  • Explainability via SHAP analysis

Insights showed that tenure and contract type reduce churn (positive influence), while high monthly or total charges increase churn risk. The model provides actionable intelligence for customer retention strategies.


๐Ÿ“ˆ GitHub Highlights

Top Languages


๐ŸŒฑ Currently Exploring

  • ๐Ÿš€ AI Agents & Knowledge-based RAG systems
  • ๐Ÿ“Š Better ways to visualize insights with Power BI & Python
  • ๐Ÿง  Research ideas in AI for accessibility and education

๐Ÿค Letโ€™s Connect

โšก Always excited to collaborate on impactful AI and Data Science projects!

Pinned Loading

  1. RadioCheck-F1-Assistant RadioCheck-F1-Assistant Public

    It is an AI-powered assistant that delivers real-time Formula 1 insights through natural conversation. Ask about driver and constructor standings, championship odds with Monte Carlo simulations, raโ€ฆ

    JavaScript

  2. Content_Creation_Agents Content_Creation_Agents Public

    AI-powered content creation framework that automates trend research, writing, and SEO. A team of specialized agents (analyst, writer, optimizer) generates high-quality, platform-specific content (bโ€ฆ

    Python

  3. Email_Optimizer_Agent Email_Optimizer_Agent Public

    Email Content Optimizer is an AI-powered app that helps you write, evaluate, and improve professional emails. It uses Groq Langchain, LangGraph agents, and the MoonshotAI Kimi K2 Instruct 0905 modeโ€ฆ

    Python 1

  4. UrduWhiz UrduWhiz Public

    UrduWhiz is an AI-powered web app that makes Urdu storybooks accessible and interactive. Users can upload scanned Urdu PDFs, which are processed with OCR and semantic search. The app summarizes, inโ€ฆ

    JavaScript

  5. PII_Detection_and_Redaction_audio_Project PII_Detection_and_Redaction_audio_Project Public

    This project focuses on detecting and redacting Personally Identifiable Information (PII) in English audio datasets. This toolkit transcribes audio, identifies sensitive entities using NLP, and redโ€ฆ

    Jupyter Notebook

  6. ASL_Image_Classification ASL_Image_Classification Public

    ASL Image Classification uses a custom Convolutional Neural Network (CNN) trained on the ASL Alphabet dataset, the app achieves ~99% validation accuracy.Users can upload their own images or select โ€ฆ

    Jupyter Notebook