Sector 01 — Energy & Renewables

Weather isn't an excuse. It's a signal.

For renewable operators, solar farms, utilities and energy traders. We combine weather forecasting systems validated against ERA5 with predictive models for generation and consumption — so schedules become more accurate, balancing penalties drop, and investment decisions rest on data, not intuition.

Discuss a project → See case studies

What hurts the sector

Three problems that cost the most.

Over two decades we worked with boards of energy operators, renewables-financing banks and farm developers. These three themes come back every time.

Balancing deviations eating into margin

Penalties for divergence between schedule and actual generation run into millions annually for an average farm. Standard weather models from brokers had sufficient accuracy a decade ago — not today.
Trading decisions on the power exchange without insight into the next 48 hours

Spot and intraday trading requires forecasts that go further than a day or two, and are probabilistic, not point-estimates. Energy traders have tools for market analysis. They rarely have tools for weather analysis.
Predictive maintenance that never got deployed

Every operator has sensors on turbines and transformers. Few have systems that extract failure-anticipating signals from those data. The gap from data to decision is the gap that costs.

Production deployments from AIGP partners

Systems in production, not slides.

The following projects come from the technology portfolio of the partner team we work with. Each operates in production, evaluated on hard metrics continuously.

Forecasting Multi-model weather forecasting system

Global 10-day forecasts outperforming Microsoft Aurora.

A fully automated system generating global forecasts every 6 hours, at 0.25° resolution and 1-hour time step, with dynamic ensembling of multiple models (ECMWF AIFS, GFS, GraphCast, ICON, GRAPES, WeatherNext) and advanced post-processing. For each of over a million grid points, model weights are independently optimized — the system adapts to the current quality of each data source.

10 days

Forecast horizon

0.25°

Spatial resolution

up to 90%

Surface pressure improvement vs reference models

every 6h

Prediction generation frequency

Analytics & decision intelligence Risk analysis system

Probabilistic answers to decision questions in <30 seconds.

A central service answering questions like "will CO2 price exceed X by end of quarter" — combining multi-agent LLMs, an automated news module, zero-knowledge reasoning and post-aggregation, delivering a probabilistic answer with reasoning.

80%

Accuracy on client validation data

<30s

Response time

~20/min

Query throughput

Funding pathways

Energy projects that fund themselves.

For the right projects, 50–80% of costs can be covered by public funding. Below is a map of programmes open or planned in the 2026–2027 horizon.

Nearest call

FENG SMART Path

Poland's flagship innovation instrument. R&D modules, innovation deployment, digitization, greening. Minimum 1M PLN, up to 140M PLN in consortium.

€7.9B budget16.06–11.08.2026

EU component

Horizon Europe · Cluster 5

Energy, climate, mobility. International research and innovation consortia, up to 100% costs.

InternationalRolling calls

Green transition

KPO — component E

Investments in grid digitization, renewables management systems, energy storage.

NationalRolling

Loans & grants

NFOŚiGW · Energy Plus

Preferential financing for renewable projects, energy efficiency and infrastructure modernization.

Grant + loanOpen

Important Not every project fits grant funding — and not every project should be forced down that path. Our first recommendation is often: "this is better done from own budget, faster." We're honest about this.

Energy sector · conversation

Have a project that makes sense? Let's look at it together.

One hour of conversation saves half a year of implementation from the wrong angle. Write — we'll set a time within two days.

Book a call →