Writing
Technical writing on knowledge graphs, graph machine learning, data engineering and AI.
Three months building a neural search and recommendation system over roughly 600,000 documents in OpenSearch. Covers why ANN covers only ~10% of your index, why filters must go inside the ANN clause, the silent recall collapse from omitting pagination_depth, the HNSW parameter ceiling chain, the copy_to lexical recall pattern, and how to tell an ML-layer 429 from a proxy 502.
Read the postBuilding a knowledge graph over Nigeria's NHMIS data for all 44 Kano LGAs, with an AI agent on top. Covers the dual-grain graph model, idempotent DHIS2 ingestion, why six Cypher templates beat one text2cypher tool, and the traversal that tells a real decline in service delivery apart from a facility that stopped filing returns.
Read the postA full technical walkthrough of the NODES 2026 talk: bipartite knowledge graph design, FastRP embeddings, graph-aware hard negative sampling via constrained random walks, stratified resampling to correct power-law skew, Random Forest on element-wise embedding products (OOB AUPRC 97.3%), and Platt re-calibration from 50% sampled prior to 5.5% population prior.
Read the postTell me what you are trying to build. I will reply within one business day.
Get in touch