Our data, in your warehouse.
The market history, grid records and weather Ask the Grid runs on, delivered to Snowflake, Databricks or your own storage for the markets and years you need.
What teams build with it
Most requests start with one question. Can we beat the day-ahead forecast at these nodes? What did congestion cost this portfolio last year? We pick the datasets around it.
Forecasting and research
Prices, load, generation and weather on one clock, with the forecast vintages a fair backtest needs.
Market analytics
Spreads, congestion and system conditions, each record traceable to the operator that published it.
Assets and portfolios
Market history joined to the generators, storage and projects it belongs to, with their location and ownership.
Data from the source. Every dataset names the publisher it came from.
ERCOT50 datasets
CAISO47 datasets
PJM46 datasets
MISO57 datasetsNYISO51 datasets
ISO-NE83 datasets
SPP59 datasetsEIA47 datasets
FERC10 datasets
NOAA6 datasets
BPA13 datasets
USGS2 datasets
Also Southern Company, ODIN, NYSERDA IEDR, HIFLD, Open-Meteo, BCA, AESO, IESO, CENACE, USACE, NOAA CPC, NWRFC, NRCS, ECCC, NIFC, TVA, NRC, FRED, OpenStreetMap, CONAGUA, CNE, HQ, PG&E, FERC, NREL ATB, CPUC, SoCalGas, and California Independent Storage.
Browse the catalogEnergy datasets
Market prices
Study spreads, price formation and the locations where power settles.
ERCOT LMP by Settlement Point
ercot.lmp_by_settlement_point
Source: ERCOT
Cadence: 5-minute (SCED)
ERCOT LMP by Bus
ercot.lmp_by_bus
Source: ERCOT
Cadence: 5-minute (SCED)
ERCOT Day-Ahead System Lambda
ercot.dam_system_lambda
Source: ERCOT
Cadence: hourly (DAM)
Open a dataset for its schema, sample and coverage. Subscription scope, history and permitted use are agreed during your evaluation.
Data provenance
A settled price, a forecast and a 60-day disclosure are different kinds of fact. Every record keeps its source, when it was published, and what it measures.
Source attached
Know the publisher, dataset and methodology behind the record.
Time made explicit
Keep observed conditions, publication times and forecast targets distinct.
Coverage you can inspect
Review history, cadence, schema and samples before choosing a package.
A defined data agreement
Agree the markets, delivery, revisions, support and permitted use your team needs.
Delivery options
Choose the datasets first. Then work with us on the history, update schedule, delivery path and use rights that fit your environment.
Your warehouse
Tell us your Snowflake or Databricks environment, cloud and region. We’ll scope the connection around your team's workflow.
Warehouse delivery planning
Your storage
Plan bulk historical access and recurring file delivery into your own lake or pipelines, with formats and update semantics agreed up front.
Bulk history + ongoing updates
Your application
Explore the existing API for targeted data access and applications. Use the developer console to inspect requests and integration examples.
Explore the developer platformQuestions about data services
If yours isn’t here, ask it in the form below.
What data can we evaluate?
Start with market prices, load and generation, weather, or asset and infrastructure records in the public catalog. Tell us the markets and analytical task you need; we’ll identify the relevant datasets and confirm the subscription scope.
Can we use our existing warehouse?
Yes—bring your existing stack into the evaluation. Tell us whether you use Snowflake, Databricks, DuckDB or another environment, along with your cloud and region. We’ll agree a delivery approach and validate the connection with you.
How much history is included?
History varies by dataset and location. Each catalog listing lets you inspect its coverage. During evaluation we confirm the required date range, known gaps and whether historical and ongoing location coverage match your needs.
How often does the data update?
Each source has its own publication clock, from market intervals to monthly filings and delayed disclosures. Catalog entries name the cadence; the data agreement defines the delivery schedule for your selected package.
Can we backtest models with it?
We’ll review the temporal requirements of your backtest, including forecast issue times and source revisions. Vintage availability differs by dataset. Latest corrected facts, historical forecasts and reanalysis serve different purposes and should be selected accordingly.
How are subscriptions priced?
Subscriptions are scoped to your datasets, markets, history, delivery and support requirements. Request an evaluation so we can define the package and its price together. Customer compute and any managed services are addressed in the delivery plan.
Can we train models or redistribute the data?
Tell us how you intend to use the data. Internal analysis, model training and customer-facing redistribution require different rights. Your data agreement sets the permitted uses for the selected sources.
What happens after we request an evaluation?
A person from our team reviews your requirements. Together we select a useful set of data, confirm coverage and rights, and plan a sample or connection against a real analytical workflow.
What are
you building?
Tell us the market, the data and the question. We’ll help shape an evaluation around the work your team needs to do.
- 01Choose a market and analytical task.
- 02Confirm coverage, history and permitted use.
- 03Plan a sample or connection in your stack.
Request a data evaluation
Your note goes directly to our team.
Loading the evaluation form…
