Distributed spatial compute
Spatial joins across billions of rows at sub-hour runtimes — in Sedona, Spark, and Databricks.
- Founded
- 2022 · Buffalo, NY
- Designation
- Veteran-owned small business
- Focus
- Geospatial data · Agentic AI · Cloud platforms
- Recent work
- Cloud geospatial · Global insurance · Critical infrastructure
Founder & Principal
Kyle Ryan
Kyle works where geospatial meets AI — pairing deep GIS practice with cloud-native data engineering and production agentic AI, a combination the market almost never delivers in one person.
That AI practice is hands-on and continuous. Kyle architects and operates production agentic systems end-to-end — multi-agent orchestration, persistent memory and retrieval architectures, model routing and fallback policy, cost governance, and secure cloud deployment — the same disciplines MapGuru brings to client systems, and the ground truth behind the counsel he gives executives on AI adoption.
The record runs through the industry's frontier. At Wherobots — the cloud-native geospatial analytics company from the creators of Apache Sedona — he served in a go-to-market solutions architecture role at the intersection of AI tooling and earth observation data. Inside global commercial insurance, he engineers the AI-assisted Databricks pipelines that ingest and harmonize natural hazard data for enterprise-scale risk understanding. Earlier, he engineered geospatial data systems across utilities, telecom, and municipal government — including the City of Palo Alto's utility GIS and chief engineer–level consulting for the Port Authority of NY & NJ.
MapGuru — a veteran-owned small business founded in 2022 in Buffalo, NY — is where that experience is put to work for clients: strategic counsel with hands still on the keyboard.
Technical depth
Where the depth shows
Cloud-native formats
Zarr, GeoParquet, COG, and STAC — worked at the format level, not just consumed.
Geodesy & CRS
Projection and datum problems root-caused at the source-code level of the open-source spatial stack.
Imagery & ML inference
Production vision-model pipelines over satellite and aerial archives, validated to numerical parity.
Agentic orchestration
Multi-agent systems, Ray and Flyte pipelines, and AI-directed spatial workflows.
Working stack
Apache Sedona · Apache Spark · Databricks · PostGIS · GDAL · Zarr · GeoParquet · STAC · H3 · PyTorch · Ray · Esri
How we work
Three commitments
Candid counsel
You get our honest read — including when the answer is “don't build this” or “AI won't help here.” Trust is the deliverable behind every deliverable.
Hands on the keyboard
Strategy and implementation from the same person. The advice is credible because we've built — recently, at scale, in environments like yours.
Built to hand off
Everything lands in your platforms, documented and owned by your team. Success is when you no longer need us — until the next hard problem.
