- Invent and productionize Transformer/RAG/Graph RAG architectures.
- Prototype and launch hybrid dense/sparse retrieval pipelines on vector DBs.
- Own high-recall NER models for multi-language text entity extraction.
- Build cross-dataset entity-resolution frameworks for data deduplication.
- Design and implement agentic workflows with robust evaluation frameworks.
- Scale ML solutions and partner with engineers for production reliability.
- Drive end-to-end project ownership from definition through deployment.
PythonMachine LearningPyTorch+2 more