- Implemented Machine Learning and NLP based approach to rule-based Entity Matching tasks, significantly improving the accuracy of matching data clusters from 52% to 68%.- Enhanced model fit and performance by fine-tuning large-scale LLM and Transformer Model for a tailored Entity Matching use-case.- Implemented TF-IDF vectorizer and compared performance on various metrics against SBERT, processing approximately 50,000+ data records, leading to insights on the efficiency of both the approaches for entity matching.- Utilized sklearn based classes including k-means and Gaussian Mixture to implement clustering algorithms to group similar entities based on cosine similarity scores.

- Implemented JavaScript to apply business rules governing master data flow processes, improving efficiency and reducing processing errors, and customized WebUI to enable more convenient access to master data across 9 teams.- Developed Change Packages that facilitated seamless data migration across 3 environments (Dev, Test, Production), reducing migration time by 5 hours/week and minimizing data loss during transition.- Planned data workflows for different teams that acted as the blueprint of the flow of customer and supplier data across systems, enabling users to create new entities and change existing ones and mapped the workflows to Web User Interface.- Created process flow for data migration from SAP MDG to Stibo STEP.