Self-directed sample, 2025

AURRA: four documents for one AI product

AURRA is a fictional AI smart thermostat I created so every document could be shown in full. It has a REST API, OAuth 2.0, webhooks and a predictive HVAC model, which means four very different readers.

For homeowners

User guide

Setup, schedules and troubleshooting in plain language, organized by task.

Open the document

For backend engineers

API reference

12+ REST endpoints with parameter schemas, error handling and Python and JavaScript samples.

Open the document

For integration engineers

Developer integration handout

OAuth 2.0 setup, telemetry endpoints, webhooks and an SDK quickstart.

Open the document

For ML and risk reviewers

Model card

Architecture, training data, feature importance, thresholds, subgroup performance and limitations, following the Google Model Card Toolkit format.

Open the document

All four documents are original. Nothing was adapted from client material.

What it shows

The same product explained four ways. Each document changes its structure, vocabulary and level of detail for its reader, from a homeowner changing a schedule to a reviewer checking how the model behaves for different groups of users.