J Williams

  • Geospatial Data Developer
  • Data Engineer
  • Research Scientist

I build the data infrastructure underneath spatial systems, where millions of records from dozens of sources have to agree before anyone can ask a question.

Selected work

All work
  • CDISAW: a queryable data infrastructure for slavery and armed conflict

    A research platform that makes 53 incompatible datasets on slavery in war answerable as one corpus. 23.5M linked event records across five ontological layers, queried through a typology-first interface, with every query citable.

    PostGISFastAPIElasticsearchH3+3
    23.5MLinked event records2026 to present
  • Topodex: contextual geocoding for conflict and human rights corpora

    A Python library that makes geocoders accountable to the document they are reading. 894M+ places resolved across nine open sources, reranked against document context, with a six-category failure taxonomy and a coherence check over everywhere a document names.

    PythonNLPGeocodingNominatim+3
    894M+Places resolved2026 to present
  • AnythingPOI: a fused, confidence-scored points of interest dataset

    OpenStreetMap and Overture Maps conflated into one deduplicated, classified, openly licensed dataset. 22.7M points across six countries, every record carrying a confidence score and the evidence behind it.

    OpenStreetMapOverture MapsH3DuckDB+3
    22.7MPoints of interest2026 to present
  • WalkGrid: personalising urban walking through environmental similarity

    A routing platform that curates walks by what makes them worth taking, from greenspace and heritage to air quality and safety, rather than by distance. 51 environmental features scored across a hexagonal grid, with route selection driven by natural language.

    H3PostGISOSRMLLM+3
    51Environmental features2020 to 2024

What I do

Pipelines

ETL that survives its author

Reproducible ingestion at volume, containerised and orchestrated, from raw source to GeoParquet, PMTiles and vector tiles.

Conflation

Entity resolution at scale

Multi-stage matching across heterogeneous sources, with validation that catches the plausible-looking wrong answer.

Spatial

Geospatial systems

PostGIS, spatial SQL, H3 discrete global grids and routing engines, where projections and joins decide correctness.

GeoAI

Representation learning

Graph neural networks over street networks, NLP geoparsing, and embeddings that make places comparable.

Contact

Let's work
together →

Open to consulting engagements and research collaborations. The fastest route is email, at james@jameswil.com.