One platform, from forward markets to single-asset dispatch. Decisions flow up from the asset. Metered data flows back down into positions, risk and settlement.
Real time, seconds
Forward, months to years
01
02
03
04
05
06
07
-
01
Physical assetsProduction, load, storage, EVs
Asset Management Master data, live telemetry
-
02
Dispatch & controlSetpoints per asset
Dispatching VPP, ATLAS EMS on site
-
03
Data & modellingTime series and formulas
Unified Data Platform Math Editor, Data Room
-
04
AnalyticsForecast, simulate, optimise
Advanced Analytics Forecasting, BI, dashboards
-
05
Physical tradingDAM, IDM, XBID, balancing
Scheduling & Bidding Automated market orders
-
06
Financial tradingForwards, PPAs, hedges
Portfolio Management One deal model
-
07
Risk & settlementPositions, risk, invoicing
Risk & Settlement Back office, customer portal
Runs as SaaS or on premises. Open through a REST API and an MCP server for AI agents.
- Assets managed
- 4,000+
- External interfaces per deployment
- 14+
- Deals per instance
- 1M+
- Availability target
- 99.99%
In production in Greece, Cyprus and Belgium with generators, suppliers, traders, aggregators and system operators.
Utilities, suppliers, aggregators and operators running on ATLAS
Built from the single asset up. Every position, forecast and contract starts from what a real asset produced, stored or consumed, at quarter-hour resolution.
Telemetry straight into the book. Output, availability and state of charge stream from each site, and market results go back as setpoints per asset.
- Field protocolsModbus TCP/RTU, OPC UA, EtherNet/IP, PROFINET, SNMP
- Grid interfaceTSO and DSO via on-site RTU, IEC 60870-5-104/101
- PlatformsHuawei FusionSolar, Fronius, AVEVA PI, Alectris
- DispatchRamp, SoC, cycling and curtailment limits checked before a bid is sent
- Solar park0.0 MW
BatterySoC 10%- €Day-ahead price€0/MWh
- PV to market
- Discharge
- Charge
- Price
Every series, one place. Asset measurements, market prices, forward curves and weather are validated on arrival and kept in one time-series model, with dashboards built on top.
- TelemetrySCADA, every 4 s
- Market pricesDay-ahead, intraday
- Forward curvesMonth, quarter, year
- WeatherThree providers, blended
| Time | Output MW | DA €/MWh | Cal-27 €/MWh | Irradiance W/m² |
|---|
Contracts are formulas over assets. PPAs, hybrid PPAs, tolling and flexibility deals are written once and reused for valuation, forecasting, risk and settlement.
// Hybrid PPA, solar + storage, per quarter-hour t
revenue[t] = shape[t, "pv", "bess"] × ppa_price
shortfall[t] = max(0, shape[t, "pv", "bess"]
− vol[t, "pv"] − vol[t, "bess"])
cashflow[t] = revenue[t] − shortfall[t] × spot[t]- Revenue
- € 19,120
- Shortfall
- € 716
- Cashflow
- € 18,404
Forecast error is a position. Each asset gets its own quarter-hour forecast, refreshed as weather and telemetry change, and every update flows into positions and bids.
- ModelsTensorFlow and Keras on weather, telemetry and prices
- ProvidersMeteologica, Meteogen and in-house models, blended and scored
- OutputUncertainty bands per quarter-hour, accuracy tracked per asset
- Forecast
- P10 to P90
- Metered
- Deviation now
- +0.4 MW
- Accuracy today
- 95.8%
- Next refresh
- 12:15
Know before it costs you. Rules run on live data and on calculations, and reach the right people the moment a threshold is crossed.
when output_mw < forecast_mw × 0.85
for 2 intervals
on wind_*
notify desk-operations- Wind farm 03 is 18% under forecast2 intervals, notified desk-operationsNow
- DAM book at 85% of its VaR limitChecked before tomorrow’s bids4 min
Trading right up to delivery. Orders go to day-ahead, intraday auctions and XBID automatically, and positions are recalculated on every fill.
- MarketsForward, DAM, IDA, XBID, ISP and RTBEM
- ContractsForwards, futures, PPAs, tolling, flexibility and capacity
Physical and financial, one book. A forced outage or forecast revision changes the hedge ratio the moment it happens, not at settlement.
- PositionsNetted per quarter-hour across physical and financial books
- MeasuresExposure, VaR, CVaR and stress tests on Monte Carlo price and volume paths
- P&LAttributed per asset, book and strategy, explained back to dispatch
- ControlsLimit checks before trades and bids go out
- Hedge ratio
- 82%
- VaR 95%, 1 day
- €4.1m
- Limit used
- 68%
Ask in plain words. The agent does the work. It reads the same governed data as your team through the ATLAS MCP server, shows its steps, and nothing reaches the market until a person approves.
Read about ATLAS and AIA portal for every producer and customer. The same data your team settles on, shown to each counterparty in their own Customer Portal.
- Sign in with a code, not a password. A one-time code arrives by email; there are no accounts to create.
- A dashboard made for them. Filtered by period or asset, downloaded as PDF, every chart exported to CSV, XLSX, JSON or PNG.
- Everything in one place. Assets, invoices, orders, contacts and documents.
- They send you data. Guided Data Panels for availability and other asset timeseries, pasted straight from a spreadsheet.
- One email, several companies. In their language, light or dark, on desktop or phone.
Solution
Start with one module, grow into the platform. Every module reads and writes the same data, so adding one never means another integration project.
Data Room
Centralized data hub for market prices, volumes, forward curves and weather data. Ingest, validate and distribute all market data across the platform in real time.
Math Editor
Built-in formula engine for custom calculations using operators, functions and logical conditions. Define pricing models, custom curves and derived signals without writing code.
Portfolio Management
Complete deal capture, position management, real-time PnL tracking and Mark-to-Market valuation. Organize deals into portfolios and books by strategy.
Risk Management
Value-at-Risk calculations, stress testing, scenario analysis and sensitivity modeling. Real-time risk exposure assessment across all positions and portfolios.
Scheduling & Bidding
Prepare and submit bids to DAM, IDM, XBID, ISP and balancing markets. Approval workflows, volume controls and fat-finger error prevention built in.
Back Office
Compliance monitoring, regulatory reporting, audit support and contract management. Centralized operations hub linking trading activity to back-office workflows.
Settlement & Invoicing
Automated settlement with shadow reconciliation against market operator statements. Invoice generation, counterparty matching and payment tracking across all settled positions.
Reporting & Dashboard
Configurable dashboards with real-time position, risk and market widgets. Comprehensive reporting suite with scheduled delivery, regulatory export formats, Power BI integration and Job Manager for automated distribution.
Forecasting
Price and volume forecasting using statistical and machine learning models. Generate forward curves, load forecasts and renewable production profiles to inform trading and bidding decisions.
Explore our Open Documentation
Workflows, integration guides and implementation resources for the ATLAS platform — open to everyone, no sign-up needed.
Introducing ATLAS EMS — Energy Management & SCADA for BESS & PV
Where ATLAS ETRM covers the commercial side — trading, risk and back office — ATLAS EMS (StellarEMS) covers the physical side: a real-time EMS & SCADA platform for battery storage and photovoltaic portfolios. Sub-second telemetry, closed-loop control and TSO/DSO grid-code compliance keep your plants optimized, safe and market-ready — working hand-in-hand with ETRM so physical operation and trading stay aligned.
Build on ATLAS — or just ask it
ATLAS is open by design: a versioned, well-documented JSON API with 280+ endpoints for your own systems — plus a native Model Context Protocol (MCP) server that puts live platform data inside your AI assistants.
ATLAS API
Extensive, documented API
A versioned JSON API spanning deals, portfolios, master data, assets, time series, invoices and the ATLASi market-operations suite — protected by JWT bearer auth, limited to 100 req/min per client, and documented live with an official Postman collection.
- 280+ endpoints across 40+ controllers
- JWT bearer tokens via POST /api/token (1-hour expiry)
- Simple JSON envelopes — requests wrap in {"data": ...}
- Versioned interactive docs & OpenAPI schema
- HTTPS-only with role-based access
Model Context Protocol
Native MCP server for AI
Connect Claude, GitHub Copilot, Cursor or any MCP-capable assistant and ask questions about your portfolio in plain language — positions, PnL, risk and forecasts answered against live ATLAS data with permissioned, auditable access.
- Native MCP server exposing 15+ domain tools & resources
- One config block works in Claude, Copilot, Cursor & more
- Read-only by default; role-based write actions
- Live platform data — no screenshots, no stale exports
- Authenticated with the platform JWT flow
$ curl -u user:pass -X POST …/api/token ·
“What’s our open risk?” → Atlas MCP answers from live data
Talk to the team behind ATLAS. Tell us about your portfolio and we will show it running on the platform.
Or write to us directly at contact@stellarblue.eu