Your questions,
answered

Dexter Energy works with automated asset-backed energy traders with wind, solar, and battery assets across European power markets.

Our customers include utilities and retailers, renewable offtakers and aggregators, independent power producers (IPPs), and prop traders.

Currently, our power and price forecasts are used by around 80 clients, collectively supporting roughly 30 GW of renewable and battery capacity.

Dexter Energy's power and price forecasts feed into the decision-making layer of a short-term power trading setup, sitting between raw market data and the algorithms or traders who act on it.

• Before the day-ahead auction, power forecasts help traders determine how much generation to bid in and at what price, minimizing the risk of ending up out of balance.
• In intraday trading, price signals and updated generation forecasts allow traders to adjust their positions as conditions change.
• At the balancing stage, real-time imbalance price forecasts help asset-backed traders and BESS optimizers decide whether to absorb risk or act.

Dexter Energy provides power and price forecasting across European short-term power markets. Coverage spans day-ahead, intraday, and balancing/imbalance markets, following the structure of each country's settlement framework.

Coverage is expanding as new markets are onboarded. If you want to check whether your market is covered for a specific product, contact us.

Dexter Energy's forecasts are built on a combination of machine learning, meteorological expertise, and proprietary data infrastructure.

Power Forecasting draws on 10+ NWP models processed through a purpose-built weather platform, plus a proprietary Solar Nowcasting model that updates every 15 minutes from satellite data. ML models are trained on asset-level production data cleaned through automated pipelines and human review.

Price Forecasting combines ML with fundamental market signals across 40+ external sources. Signals are market-impact aware, meaning they account for the effect that acting on a forecast has on the market itself.

Dexter Energy’s power and price forecasts are built from the ground up for automated energy trading.

Our technology and data were designed specifically for the way traders work today. Our team has hands-on trading roots, so we think in P&L and settlement terms, not just meteorology. Furthermore, our signals are market-impact aware and enriched by a proprietary data moat that other forecasters can't easily replicate.

Dexter Energy's Power Forecasting covers wind, solar, and prosumption assets across Europe.

Dexter Energy's power forecasts are produced at the asset level: each wind park, solar park, or prosumption connection is forecast individually.

Assets belonging to the same customer within a bidding zone are grouped into a cluster, and clusters are grouped into a portfolio.

Dexter Energy's Wind and Solar Power Forecasts consistently rank in the top 3 in independent forecasting trials across Europe.

When evaluating forecast quality, the metrics that matter are overall accuracy (NMAE / wMAPE), performance in extreme situations (RMSE), and bias. Accuracy varies by asset, portfolio, and market conditions; there's no single number that applies universally. We explain our recommended approach and reasoning in this white paper on evaluating power forecasts.

A more accurate forecast means fewer and smaller deviations between what you bid and what you produce, which reduces your exposure to imbalance settlement prices. Customers using Dexter's forecasts have reduced balancing costs by up to 35%.

Dexter Energy draws on 10+ traditional NWP and AI models, combining global and high-resolution local sources to achieve better accuracy than any single model.

We store over a petabyte of historical weather forecast data and ingest over one terabyte of new data daily from all major weather providers. We continuously benchmark models to identify which inputs perform best in which conditions. Our infrastructure took years to build and is a significant part of what makes our forecasts hard to replicate.

For solar, we also run a proprietary Nowcasting model fed by satellite data, giving traders accurate irradiance predictions for the crucial 0-3 hour window ahead.

Dexter Energy's Wind Power Forecasts update every hour. Our Solar Forecast updates every 15 minutes thanks to our nowcasting model. The Prosumption Forecast updates four times per day.

All power forecasts are delivered at 15-minute resolution, covering today and the next three days. This can be extended to up to 10 days ahead, on request.

Yes. Dexter's Wind Power Forecast accounts for high-wind shutdown and wind turbine icing.

High-wind shutdown is supported for all wind assets across Europe. When forecast wind speeds approach a turbine's cut-out threshold, the forecast is gradually ramped down to reflect the expected loss of production.

Icing coverage is available in select markets. When ice build-up is forecast to reduce or halt turbine output, this is reflected in the forecast.

Customers can also opt in to Power Loss Alert emails, which notify them ahead of extreme weather events and include a breakdown of expected production losses, helping them adjust trading positions in advance.

More power loss features are in development.

Yes. Please get in touch to discuss your portfolio’s needs.

Dexter Energy delivers all forecasts via a self-serve RESTful API, designed to integrate directly into trading algorithms or execution platforms. However, we are open to custom integrations into FTP servers and customer APIs as well.

For Power Forecasting, customers can independently add or remove assets without our involvement. Forecasts are available in real time as soon as each run completes.

The API documentation is available here.

Dexter offers a structured Proof of Value (POV) trial before you commit.

For Power Forecasting, the trial follows a two-phase process: onboarding, which covers data upload, validation, and the creation of a staging environment, followed by a live trial of typically 3 months with regular check-ins and ongoing performance and data quality monitoring.

At the end of the trial, Dexter provides a full evaluation including accuracy metrics (NMAE, balancing costs, balancing volume) and a business case showing the financial impact of the forecast on your portfolio.

For Price Forecasting, trial options are available; get in touch for details.

Getting started with Dexter’s Power Forecasting requires asset location, historical production data, and basic operational data.

More data generally means a better-performing model. For wind and solar, we ask for at least 2 years of historical production data, with 3 years preferred. For newly built assets with no history, we can still run forecasts, though accuracy improves as actuals accumulate.

Data quality matters as much as quantity. During onboarding, Dexter reviews and cleans incoming production data — flagging curtailment periods, maintenance outages, and measurement errors — so you don't need to deliver perfect data from the start.

Dexter's pricing is structured as a SaaS subscription.

The subscription fee depends on the number of countries or markets you need covered, and the size of your portfolio (installed capacity or number of assets). This means pricing reflects the actual scope of what you're forecasting, rather than a flat, one-size-fits-all rate.

For an exact quote based on your setup, get in touch, and our team will walk you through the details.

Dexter Energy uses a combination of automated monitoring and human review to maintain forecast quality over time.

On the automated side, we continuously track performance metrics at the asset and portfolio level, monitor for data drift, and run proactive checks on incoming customer data.

When automation flags an issue or nuance is required, a human steps in as part of our human-in-the-loop approach.

Models are retrained regularly as new data comes in, and any code changes are validated against historical performance before going live.

Dexter Energy's API and forecast infrastructure operate at over 99.8% uptime.

Reliability is built into our product architecture. Forecasts include automatic retries, fallback runs, and continuous uptime monitoring. If a forecast run fails, customers can always retrieve the most recent successful run via the API. A dedicated on-call team monitors alerts during working hours and acts on any issues before they affect trading.

Models are continuously refreshed as new data comes in, with no manual steps or downtime required on the customer side.

After onboarding, every Dexter customer has a dedicated customer service manager and access to ongoing performance and data quality monitoring.

Support includes regular check-ins, proactive alerts when data issues are detected, and asset- and portfolio-level performance reviews. Dexter takes a partner approach rather than a hands-off vendor model; where possible, we flag issues before customers do.

If there is an issue with a forecast, customers can always retrieve the most recent successful run via the API as a fallback. The on-call team monitors alerts during working hours and investigates promptly.