Software Engineer, Geospatial Systems

Paces•New York, NY

2d•$150,000 - $220,000•Hybrid

About The Position

Paces is building the development engine for renewable infrastructure and power projects. We have built Agentic AI that automates early-stage due diligence and pairs it with in-house expertise to accelerate late-stage diligence. The result: projects move from siting to construction-ready in a fraction of the time. Our platform streamlines siting, interconnection, and permitting for renewable energy and data center developers, real estate firms, EV charging companies, and more. We are looking for exceptional talents to join our growing Software Engineering team. You will be working closely with power and permitting industry experts to automate power development!

Requirements

4+ years of experience building data intensive software in a similar role
Strong computer-science fundamentals: data structures, algorithms, graph theory, and a feel for where time and memory go.
Fluency in a systems/backend language.
A track record of building performant, correct backend systems — you can reason about a hot path and read a query plan.
Experience with DuckDB, Apache Arrow, or comparable columnar/analytical data systems.
A research mindset: comfort with open-ended problems, prototyping to learn, and knowing when to harden a prototype into a production system.
Comfort owning ambiguous problems end-to-end and communicating tradeoffs clearly in writing.

Nice To Haves

Experience with a systems language — Rust, C, C++, Zig, or similar.
A geospatial / GIS background: PostGIS, GDAL/raster drivers, GeoArrow, projections and coordinate systems, or computational-geometry libraries (GEOS, georust/geo).
ML, embeddings, or geospatial foundation models (e.g. Clay).
Familiarity with zarr / GeoZarr or other large-scale array/raster formats.
Large-scale spatial pipelines, spatial indexing, or distributed/ephemeral compute.
Open source contributor.

Responsibilities

Research, prototype, and ship core parts of the geospatial engine at the heart of Paces — large-scale spatial search and analysis that energy developers depend on.
Take open-ended, hard problems in computational geometry and performance — over tens of millions of parcels and nationwide infrastructure data — from prototype to production.
Work across the full spatial stack: DuckDB, Arrow/GeoArrow, file-based spatial indexing, and a OLTP-backed data lake.
Explore the frontier of geospatial computation — for example geospatial embeddings and foundation models (e.g. Clay) — and decide what's worth bringing into the product.
Set the bar for correctness and performance on a system where wrong or slow answers carry real cost.