Lokesh Sharma

Built an agile ontology engineering practice, cutting data-modeling time from ~2 months to 3 weeks. Designed modular ontology architecture and led KG construction end to end, holding accountability for 8+ customer instances.

Filed 3 patents and co-authored research papers on Temporal Knowledge Graph forecasting. Built entity-linking data pipelines and developed AI models, including GNNs, with dockerized inference.

Integrated labeled property graph schemas into Azure data ingestion pipelines, improving shipment-tracking/query performance on 10M+ records. Extended patent-document extraction pipeline with NLP and entity linking.

Benchmarked state-of-the-art GNN architectures for TKG forecasting on 5 large datasets through systematic experimentation tracked in TensorBoard.

Built ETL pipelines consolidating various data sources to automate manual workflows and generated descriptive analytics reports. Developed RESTful APIs and interactive BI dashboards for 800+ sales representatives.

Built and deployed classification and clustering models forecasting customer churn and portfolio financial risk. Delivered data-driven M&A deal analytics including sampling design, A/B testing, and executive dashboards.

Completed GraphDB Knowledge Graph Engineer certification in February 2025.

Completed Ontotext GraphDB Developer certification in February 2025.

Completed Neo4j Certified Professional certification in November 2023.

M.Sc. in Web and Data Science with coursework in Network Theory, Big Data, AI, Machine Learning, and Semantic Web.

B.E. in Instrumentation Technology with coursework in Microprocessors, Embedded Systems, Neural Networks, and Arduino.