This AI-powered Auto Workshop platform was developed in collaboration with Carfix.pk to automate vehicle fault detection and repair cost estimation. By integrating NLP, Computer Vision, and Predictive Analytics, the system analyzes text, images, and sound data from vehicles to deliver accurate diagnostics and recommendations in real time.

🔑 Key Features:

✅ Fault Diagnosis Automation – Identifies common and complex vehicle issues using AI models.
✅ Repair Cost Estimation – Predicts repair expenses with high accuracy using predictive modeling.
✅ Multi-Modal Inputs – Processes text (symptoms), images (YOLOv8-based car part detection), and audio (Librosa for sound anomaly detection).
✅ Interactive Chatbot – NLP-driven assistant to guide users through problem reporting and solution options.

🛠️ Technologies Used:

• Python | NLP | YOLOv8 | Librosa | TF-IDF | Scikit-learn | Flask | FastAPI | SQL | AWS
  
  

🚀 Impact: Helped streamline customer support and reduce manual inspection time, enabling faster and more transparent repair services.

Bringing Conversations to Life with AI & NLP

This AI-powered voice chatbot leverages Groq’s Distil-Whisper and Llama 3 models to provide real-time speech-to-text transcription and intelligent responses. Designed for seamless, human-like interactions, the chatbot enhances user engagement by understanding and responding accurately in natural language.

🛠️ Key Features:

✅ Real-Time Speech Recognition – Converts spoken words into text instantly.
✅ AI-Powered Responses – Uses Llama 3 for context-aware replies.
✅ User-Friendly Interface – Clean, intuitive chatbot UI for better accessibility.
✅ Deployed on Hugging Face Spaces – Easily accessible and scalable for users.

🔗 Technologies Used:

🔹 Python | NLP | Groq API | Distil-Whisper | Llama 3 | Hugging Face | Gradio

Predicting Future Crop Export Values with AI & Machine Learning

This project focused on forecasting crop export values for the next three years using historical data from multiple countries. By implementing a Multilayer Perceptron (MLP) neural network, I ensured accurate predictions while preventing overfitting through advanced machine learning techniques.

🛠️ Key Actions & Solutions:

✅ Data Cleaning & Preprocessing – Imputed missing values, normalized datasets, and engineered features.
✅ Exploratory Data Analysis (EDA) – Identified trends and correlations affecting export values.
✅ MLP Model Implementation – Designed a neural network with optimized activation functions and dropout layers.
✅ Overfitting Prevention – Used train-test split, early stopping, and regularization to enhance model generalization.
✅ Performance Evaluation – Assessed model accuracy using Mean Squared Error (MSE).

🔗 Technologies Used:

🔹 Python | TensorFlow | Keras | Pandas | Scikit-learn | MLP Neural Networks

Using Data Science to Enhance Market Insights & Sales Performance

This project focused on analyzing sales disparities and customer preferences in Queensland compared to other regions (Victoria & New South Wales). Through market segmentation, customer insights, and competitive analysis, I developed data-driven strategies to improve sales and optimize marketing campaigns.

🛠️ Key Actions & Solutions:

✅ Market Segmentation Analysis – Categorized customers by geography, demographics, and behavior.
✅ Competitive Analysis – Analyzed competitor pricing, product differentiation, and market positioning.
✅ Customer Insights & Preferences – Identified key trends influencing purchasing decisions.
✅ Business Intelligence Dashboard – Created Power BI dashboards for real-time monitoring of sales & performance.
✅ Targeted Marketing Strategies – Developed personalized advertising campaigns based on customer behavior.

🔗 Technologies Used:

🔹 Power BI | Python | SQL | Market Analytics | Business Intelligence | Customer Segmentation

Enhancing Data-Driven Decision-Making in Supply Chain Management

This project involved the development of a dynamic Power BI dashboard to optimize Unilever's supply chain operations. The dashboard provides a real-time overview of revenue, order quantity, stock levels, defect rates, and shipping costs, enabling efficient decision-making and process optimization.

🛠️ Key Actions & Solutions:

✅ Average Lead Time Analysis – Evaluated transportation modes (Air, Road, Rail, Sea) to reduce shipping delays.
✅ Stock & Inventory Management – Tracked stock levels for different product categories (Skincare, Haircare, Cosmetics).
✅ Revenue & Sales Analysis – Identified top-performing products and revenue distribution across demographics.
✅ Supplier Performance Monitoring – Assessed defect rates per supplier to improve quality control.
✅ Shipping Cost Optimization – Analyzed shipping carriers to reduce cost per order and enhance logistics.

🔗 Technologies Used:

🔹 Power BI | SQL | Data Visualization | Supply Chain Analytics | Business Intelligence