Skip to content
NordShifts

Energy

AI agents that work from tested models, not guesses

Each agent is Claude working with the platform's own models through read-only tools. The models do the numbers; the agent plans, explains and drafts, and a person approves anything that matters.

01

Numbers come from models

Every figure is computed by a tested optimiser, forecast or audit and returned by a tool. The agent may not invent or estimate one.

02

Every figure is cited

Each number in an answer names its data source and date, so an engineer or a lender can trace it back.

03

A person approves

Plans and bids are drafts. They are checked against the plant's limits and wait for approval; nothing is sent to any market.

Six agents

For district heating control rooms

Heat dispatcher

Builds tomorrow's plan for the whole plant and drafts the reserve bids the electric boiler can offer without risking heat supply.

  1. 1Forecasts tomorrow's heat demand with a range, from the plant's own meter history
  2. 2Plans CHP, boilers, heat pump and tank against tomorrow's prices
  3. 3Drafts reserve bids within the boiler's headroom and flags uncertain hours
  4. 4Waits for the shift lead to approve or reject

For battery owners and lenders

Revenue auditor

Checks what a battery earned against the same battery following the frozen public forecast and with hindsight, month by month.

  1. 1Reads the asset's dispatch export
  2. 2Prices it at the day-ahead auction
  3. 3Compares with the frozen-forecast and hindsight benchmarks
  4. 4Writes the lender report with ledger fingerprints, as PDF

For analysts, developers and investors

Market analyst

Answers questions about what batteries and flexible assets earned, by zone, month and revenue stream, with a table and its source.

  1. 1Reads monthly arbitrage and reserve value per MW
  2. 2Adds the imbalance nowcast and price statistics
  3. 3Answers with a table, the drivers and the caveats

For installers and heat consultants

Site screener

Runs the battery or heat case on a customer's meter file and drafts the proposal for their manager.

  1. 1Reads the customer's meter file and tariff
  2. 2Runs the battery case or the heat case
  3. 3Drafts the proposal with savings, investment and payback, as PDF

For investment committees and lenders

Investment analyst

Writes an investment memo on a battery, electric boiler, heat pump or heat storage in a chosen market, from the platform's data.

  1. 1Values the asset's revenue streams in the zone
  2. 2Ranks the market among 25 in Europe, Japan and Australia
  3. 3Values heat assets against a whole plant
  4. 4Writes the memo with risks and what to verify

For heat utilities' analysts

Model selector

Recommends which demand forecasting model to use and which settings to try, from the scored model catalogue and the plant's own data.

  1. 1Reads every model's honest day-ahead scores
  2. 2Profiles the plant's heat-meter file
  3. 3Recommends a model and a runner-up, with reasons
  4. 4Proposes setting changes worth testing

Try an agent on your own data

Send a year of heat-meter or battery data and we run the agents on it with you.

Ask for a pilot
AI agents for heat plants, batteries and lenders