Peer-reviewed research, white papers, and technical writing spanning AI engineering, knowledge graphs, and applied machine learning.
Conference papers and proceedings with formal review
Research on applying knowledge graph techniques to enable recommendation systems and trend analysis within corporate research communities. Accepted at the premier semantic web conference.
Technical reports and architecture documentation
White paper detailing the infrastructure architecture and engineering challenges behind IBM's generative AI model development pipeline, covering distributed training, model serving, and operational tooling at scale.
Articles and tutorials published on IBM Developer
Technical deep-dive on building a production-grade entity grouping pipeline combining semantic embeddings and K-means clustering. Covers end-to-end architecture: data preparation, profile construction, embedding generation via Watsonx/IBM Granite, K-means implementation, cluster interpretation, and multi-tier caching. Derived from production work on the External Recognition platform's research community discovery system (405 researchers → 5 semantic communities using Llama 3.3 70B 4096-dimensional embeddings).
Published article on IBM Developer covering structured output techniques for large language models using JSON-based prompting strategies. Explores patterns for enforcing deterministic, machine-readable LLM outputs in production systems.
Step-by-step tutorial on IBM Developer for building a retrieval-augmented generation system over Markdown documentation. Covers document ingestion, chunking strategies, vector embeddings, and conversational retrieval - the technical basis for the NEDB 2026 poster on Conversational RAG for Markdown Docs.