An end-to-end decision optimization platform for electrical power assets

We help utilities and power asset managers leverage quantitative decision-making and AI to make the most of their assets.

Leverage your data in a unified environment

Optimize and automate your decisions

Maximize your profits

L'application

What we do

Your assets

Renewable Generation

Flexible Generation

Storage

Datasets we integrate

Market

Weather

Grid

Sensors

Your decisions

Portfolio Intelligence:

Production optimisation:

Trading:

Asset analytics and maintenance optimization

Power production planning optimization

Algorithmic power trading strategies

Demand and Production Forecasting

Flexibility optimization

Imbalance management

SVD Power Platform ©

Fonctions

Capabilities

Power Portfolio Digital Twin

We develop digital twins of a large multi-asset portfolio by integrating all relevant internal and external data sources in our cloud-based platform.

Leveraging an open and flexible object-based data model, we model each component of your portfolio with their attributes, time series and links to other objects.

Multiple built-in portfolio intelligence capabilities are provided, ranging from simple analytics and visualizations to help track marginal productivity or unplanned downtimes KPIs, to AI-based demand and production forecasting tools.

Production Planning Optimization

Our platform incorporates production planning features to help manage flexibility in your portfolio (battery storage, flexible power generation, demand flexibility...)

We develop custom mathematical unit commitment models, representing your unique supply-demand equilibrium problem and asset-specific constraints.

We then provide advanced algorithms to solve them efficiently at a high frequency, outputting an optimized forward power production schedule.

Algorithmic Power Trading

We connect your power assets to the European wholesale spot markets, from day-ahead to intraday, allowing automation of trading operations.

We provide capabilities to automate your existing power trading logic for various use cases: power sales, storage, hedging, imbalance management...

We also develop and backtest customised AI-powered algorithmic strategies to optimize open positions based on forecast deviations, and trade your asset's production into the right market at the best possible price.

Use cases examples

/01

Production planning for a utility operating a mix of renewable and flexible generation assets and selling directly to industrials.

Challenges:

Customer demand is fluctuating and needs to be forecasted to plan production.
Production capabilities depend on complex multivariate constraints (weather, marginal cost, downtimes...).
Production planning is a complex stochastic optimization problem known as Unit Commitment, requiring specialised skills to model and solve efficiently.
Deploying a custom optimisation software requires a team composed of software engineers, data engineers, data scientists and optimisation scientists.

Solution:

Integrate the different relevant data sources in our unified environment: past demand, past production, costs, grid, weather, sensors, market data...
Leverage our time series forecasting AI models to generate consumer demand forecasts.
Leverage our Mathematical Optimisation models for Production Planning integrating the various datasets and forecasts as constraints and objective functions to produce an optimised production schedule.

/02

PPA imbalance management for an asset manager exploiting a renewable generation portfolio.

Challenges:

As part of a PPA contract, the asset manager needs to provide a fixed power quantity to its customers. Imbalance can be created when the actual production doesn't meet the terms of the contract, due to weather for example.
The asset manager has no in-house trading desk: imbalance management needs to be outsourced to a third party trading house. This creates latency and a lack of transparency in the trading decisions.
This can lead to suboptimal trading decisions due to power price volatility.

Solution:

Generate AI-based power production forecasts to manage imbalance proactively.
Automate trading execution using our APIs to major power exchanges such as EPEX. In-house analysis of costs incurred with our visualization capabilities.
We also provide statistical models for power spot price forecasting, in order to optimize the trading decisions. Those can be leveraged to create a custom trading strategy. We provide backtesting tools to analyze its performance.

À propos

Our story

We are a team of AI and optimisation experts, with experience solving complex problems for some of the largest energy companies in the world.

Our conviction is that as energy production becomes more fragmented, energy grids more complex, and power markets more volatile, the complexity of operating a portfolio of power assets will only increase.

Managing this complexity will require processing increasingly large amounts of data from disparate internal and external sources,requiring in turn software engineering and AI expertise.

Our mission is to help power production companies of all sizes master data-driven decision-making to maximise their profits.

Avis

The team

Antoine Bambade, Co-founder

Antoine holds a PhD in Artificial Intelligence from INRIA.
Antoine also holds a M.S. in Mathematical Statistics from Cambridge University and M.S. in Engineering from Ecole Polytechnique.
Antoine started his career as an Optimization Researcher at EDF, the leading French power utility. He focused on problems such as grid supply-demand equilibrium or optimal maintenance scheduling.

Victor Catteau, Co-founder

Victor holds a M.S. in Operations Research from Columbia University and an M.S. in Engineering from Ecole Polytechnique.
He started his career as a quantitative researcher at J.P. Morgan, focusing on applications of AI to algorithmic trading and investment.
He then worked as a Machine Learning Engineer at C3.ai, a San-Francisco based tech company developing AI software, leading projects for customers in the energy sector, such as Shell, Engie or Fortum.
He started a first software company backed by Entrepreneur First.

Companies we worked with:

Contact