Insights

Research, methodology, and field notes from the programs we work on.

From the people building the platform with utilities and grid operators.

How the numbers get produced

For technical teams evaluating the methodology behind the data.

Data & Methodology

GRETA Documentation

Technical documentation for GRETA (Generator Real-Time Emissions Assignment), Singularity’s generator-level emissions model deployed at MISO.

Download PDF →
Data & Methodology

CarbonFlow Documentation

Technical documentation for CarbonFlow, Singularity’s framework for tracing carbon emissions from generation sources through to consumption.

Download PDF →
Data & Methodology

Validating Customer Data at Singularity Energy

How we ensure the quality and integrity of the utility data that powers clean energy reporting: the validation logic, edge cases, and tradeoffs involved in working with real-world interval data.

Data & Methodology

Evaluating the Hourly Emissions Intensity of the US Electricity System

Peer-reviewed research on hourly emissions intensity methodology across US grid regions, published in Environmental Research Letters.

Read paper →
Data & Methodology

Hourly Accounting of Carbon Emissions from Electricity Consumption

The foundational methodology paper on hourly carbon accounting for electricity consumers, published in Environmental Research Letters.

Read paper →

Curious how we build this?

For engineers considering Singularity: how our team works, and what we’ve learned building it. Full archive →

Engineering

Leverage this AWS feature to securely share S3 files with your users

How to use AWS pre-signed URLs to give users secure, time-limited access to files without exposing your bucket publicly or requiring AWS credentials.

Engineering

How to Use Sphinx to Document a Python Project Spread over Several Repositories

A practical walkthrough for setting up Sphinx documentation across a multi-repo Python project, with cross-referencing and consistent theming.

Engineering

How Not To Write An AI SQL Generator

Hard-won lessons from building an AI-powered SQL generation tool: the patterns that failed, the ones that worked, and what to avoid.

Ready to see it on a real program?

One system of record, built to run the program. We’ll walk through how Singularity applies to your program structure. 30 minutes, focused on your program.

Get a demo →

Large C&I buyer? If you want hourly clean energy data from your utility, ask them about the platform. The Georgia Power deployment with Google is an example of what that looks like.