Discovery & catalogue
Find historised objects, classify them semantically and review uncertain matches together.
Open-source project / ioBroker
AI
Analytics.
ioBroker.ai-analytics
Understand Smart Home data instead of merely storing it.
The ioBroker adapter combines historical data with explainable AI checks: it discovers relevant objects, answers questions about consumption and explains unusual observations.
View repository on GitHub ↗The idea
Analytics that do not only report an anomaly, but show which data supports it and why it matters.
01 / Approach
History becomes context.
The adapter uses existing History data from History, InfluxDB or SQL. It does not enable logging or change foreign ioBroker objects.
A semantic catalog provides the foundation for natural-language questions, period comparisons and proactive checks. A statistical pre-analysis filters unusual candidates before a language model is asked for an explanation.
View the project on GitHub ↗02 / Features
Built for
real data.
Natural-language questions
Compare consumption, device usage and periods with type-aware calculations.
Explainable anomalies
Pre-analyse baselines, trends, outliers and data gaps and explain them clearly.
Flexible and privacy-aware
Use OpenAI, Anthropic, OpenRouter or local OpenAI-compatible models such as Ollama.
03 / Status
In development.
Beta.
Ready to try: Download ioBroker.ai-analytics, discover your Smart Home data in a new way and help make the project better with your feedback. Its core is MIT-licensed.
Download now ↗Contact ↗