Automated the end-to-end TPRM process, including inherent assessment, residual assessment, and vendor response validation, ensuring seamless risk identification, control mapping, and compliance validation.

Integrated CrewAI tools like FileReadTool for data extraction, FirecrawlSearchTool and FirecrawlScrapeWebsiteTool for web-based compliance validation, and RagTool for real-time structured and unstructured data analysis, enabling accurate decision-making.

Built a scalable framework with QA agents to ensure quality assurance and consistency across workflows.

Generated and processed a 9.75M token dataset from 15 Indian law textbooks for continual pretraining.

Engineered a LoRA continual pretraining for LLaMA3 8B on a custom dataset using A100 GPUs for 20+ hours.

Built a chatbot that processes and analyzes user-uploaded documents or folders using Claude 3.5 to deliver insights with high accuracy and contextual understanding.

Developed an end-to-end customer engagement system, integrating AWS services for SMS, WhatsApp, email, and IVR channels.

Implemented a callback tracking system across communication channels.

Conducted load testing to optimize the architecture, resulting in enhanced system performance.

Configured Power BI reports to gain insights into customer engagement, guiding data-driven communication strategy improvements.

Improved responsiveness by 30% for a superior user experience.

Conducted data cleaning, feature selection, and analysis for predicting employee attrition using Pandas.

Conducted exploratory data analysis on a dataset with 41K+ samples using Pandas and Matplotlib.

Predicted term deposit subscription of a bank marketing campaign.

Applied imputation methods and tested Logistic Regression and Decision Tree models, achieving accuracies of 74% and 85%.

Addressed class imbalance with oversampling via imblearn and achieved 88% accuracy with an XGBoost classifier.