- Worked under Prof. Phanikumar Gandham on developing a Natural Language Interface using Large Language Models to access highly connected ontology data.
- Implemented a pipeline to access ontology for material science data using Apache Jena Fuseki Server.
- Developed a proof of concept NLP pipeline with Retrieval Augmented Generation based on LangChain and Gemini API.
- Created system to translate Natural Language queries into SPARQL queries for Knowledge Graph exploration.
- Implemented various thought and reasoning frameworks for Large Language Models to improve accuracy by 35%.
- Demonstrated that Ontologies with Natural Language Interfaces provide 50% better inference than Tabular Data.
- Demonstrated the potential for democratizing data access through Natural Language interfaces to Knowledge Graphs.
