Projects

Built an interactive NLP platform handling inputs up to 900+ words, delivering sentiment, emotion and summary insights in under 7 seconds using Gen AI.

Integrated Hugging Face Transformers and Pipelines to perform 3 key NLP tasks using pre-trained models fine-tuned on benchmark datasets.

Built 4+ interactive visualizations (word frequency charts, sentiment distribution, and keyword analysis) to improve data interpretation and user decision-making.

Designed and normalized a hospital database by segregating 6+ entities (Departments, Doctors, Patients, Appointments, Reports, Prescriptions, Bills) from Excel into structured relational tables. 

Implemented 1 trigger and 2 parameterized stored procedures to enforce appointment validation and enable role-based data access and monthly revenue reporting. 

Improved data integrity and scalability by eliminating 100% duplicate and past-date appointment entries through automated validation logic in MS SQL. 

Built an end-to-end analytics pipeline processing 100K+ transactional records, enabling analysis of $10.8M in revenue and identifying top product categories contributing 26% of total sales.

Cleaned and transformed 80K+ orders, 75K+ customers, and 32K+ products using Python (Pandas, Regex) and SQL, improving data accuracy and ensuring reliable business performance analysis.

Developed an interactive Power BI dashboard with 10+ KPIs, tracking $10.8M sales and 80K+ orders, and uncovered delivery delays associated with 38% lower review scores, highlighting operational improvement opportunities.