Naveed Afzal
I’m a Machine Learning Research Engineer based in Tokyo, specializing in Natural Language Processing. My work focuses on structure extraction from documents and converting unstructured information into machine-readable representations, including knowledge graphs.
Current Focus
At Stockmark Inc., I'm working on semantic metadata extraction for Retrieval-Augmented Generation (RAG) and its integration within an agentic framework to improve enterprise information workflows.
Previous Work
My past projects include building a manufacturing-domain knowledge graph, fine-tuning language models for entity–relationship extraction, and developing dialogue understanding systems from conversational transcripts.
Talks and Publications
- • Talk: LLM/VLM Business Applications - [Event URL] • [Slides]
- • Poster Presentation at YANS 2024 - [Blog Post]
- • Modularity Optimization as a Training Criterion for Graph Neural Networks, CompleNet 2018 - [Paper] • [Code]
Projects
I'm also working on Japan AI List , a searchable database of AI companies, products and services in the Japanese Market. It has also been featured on betalist.
Skills & Technologies
Here are some of the technologies I've been working with:
Latest Blog Posts
Here are some of my latest blog posts. You can find more on my blog page.