RAG Integration with LLMOps

• Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline with NLP-driven document querying using an LLM.
• Integrated LLMOps best practices — prompt versioning, model monitoring, and evaluation metrics — for reliable, reproducible AI outputs.
• Applied embedding optimisation and chunking strategies to improve retrieval accuracy and response quality.
• Built an interactive Streamlit interface for natural language querying, reducing manual search time significantly.

Credit Risk Prediction

• Developed a Random Forest credit risk classifier achieving 78.5% accuracy and deployed it using Streamlit.
• Performed data preprocessing, feature engineering, and model evaluation to improve prediction accuracy.
• Deployed the model as an interactive web application for real-time credit risk assessment.

E-Commerce Analysis System

• Analyzed e-commerce sales data using Python and SQL to uncover trends in customer behavior, revenue, and product performance.
• Built interactive Power BI dashboards to visualize key business metrics and support data-driven decision-making.
• Performed data cleaning, aggregation, and querying using SQL to prepare datasets for analysis and reporting.

B.Sc. — Mathematics, Statistics & Data Science

Higher Secondary Certificate (HSC)

Secondary School Certificate (SSC)

Python Course Certification

AI SPARKS: Ideas into Impact

AI Careers for Women