NOAA Technology Partnerships Office · May 2026
NOAA AOML and Silurian signed a CRADA to develop AI-driven tropical cyclone forecasting, build an ML-ready cyclone database, and improve the GFT model's tracking and intensity prediction.
Foundation models to simulate Earth, starting with weather.
Company research
Public company, workplace, funding, and market signals
Updated Jul 29, 2026
Overview
Silurian AI builds foundation models to simulate the Earth system, starting with weather, and sells AI-driven forecasting products for energy, utilities, infrastructure, and government use cases.
Primary product
GFT (Generative Forecasting Transformer) weather / Earth forecasting API
Founded
2024
Headquarters
Kirkland, Washington, United States
Team size
10-20
Industry
Information Services
Sub-industry
Earth intelligence / weather forecasting foundation models
Offices
Kirkland, Washington, United States
Seattle, United States
London, UK
Business model
API
enterprise deployments
B2B forecasting software
Funding
Stage
pre-seed
Total raised
$500K
Latest round
Convertible Note · Sep 2024
Latest amount
$500K
Sep 2024 · Y Combinator
People & investors
Leadership
Chief Executive Officer / Co-Founder
Scientific Advisor / Co-Founder
Chief Engineer / Co-Founder
Chief Business Officer
Chief Strategy Officer
Founding Research Scientist
Founding Research Engineer
Founders
Co-Founder
Work & culture
Research-heavy, mission-driven, and small-team oriented, with a strong emphasis on scientists and engineers building frontier AI for the physical world and deploying it into high-stakes operational settings.
Visa sponsorship
No
Compensation
Public YC job posting listed $120K-$250K base pay and 0.25%-1.0% equity for a Machine Learning Research Engineer / Scientist role; public benefit details were not surfaced.
Product & market
Pricing
Public pricing is not disclosed; the company appears to sell via API access and enterprise deployment / demo-led sales.
Differentiators
- —Foundation models for the Earth system rather than purely physics-based forecasting
- —High-resolution, rapidly refreshing regional forecasts
- —Asset- and infrastructure-level risk use cases such as rime ice and grid operations
- —Operational API for tropical cyclones and weather applications
- —Post-training on domain-specific sensor and utility data
- —Public benchmark claims against ECMWF, NOAA, and related systems
Technology
Python
TypeScript
Rust
JAX
PyTorch
Node.js
Vercel
Cloudflare Workers
Deno
Bun
React Native
NEXRAD radar data
APIs
foundation models
Generative Forecasting Transformer (GFT)
GFT-C
GFT-US
Customers
Hydro-Québec
Disaster Tech
energy organizations
government agencies
utilities / grid operators
Competitors
ECMWF
NOAA / NHC
The Weather Company
Tomorrow.io
Meteomatics
Atmo
Scale & contact
Estimated revenue
about $0.5M-$2.5M ARR (public estimates cluster around ~$1M)
Estimated monthly visits
3.9K
Traffic estimate as of Jun 2026
Latest news
Silurian AI · Oct 2025
Silurian describes Hydro-Québec-backed GFT-HQ results for day-ahead rime-ice alerts, reporting 0.73 average precision versus a 0.37 climatology baseline and use cases in grid operations.
Silurian AI · Oct 2025
Silurian details how post-training GFT on utility sensor data improves power-grid forecasting metrics, including lower MAE for temperature, wind speed, and precipitation.
Silurian AI · Jun 2025
Silurian launched GFT-C as an operational hurricane forecasting API and announced an initial integration with Disaster Tech's PRATUS decision-support platform.
Silurian AI · May 2025
Silurian announced the Nature publication of Aurora and positioned it as the scientific basis for its GFT family, including weather, air quality, wave, and cyclone forecasting.
Jobs and careers at Silurian
0 jobs at Silurian
No open positions
Check back later for new openings