Roninix
AI & Home Automation Lab
We build and test AI systems — from benchmarking language models to deploying autonomous agents — and apply what we learn to a fully automated home powered by Home Assistant.
Explore the projectsAI Model Testing
We continuously evaluate and benchmark large language models across reasoning, coding, and agentic tasks, comparing real-world performance beyond published leaderboards. The goal: understand what each model is actually good at, not just what it claims to be good at.
AI Agent
Our AI agent project explores autonomous, tool-using systems that can plan, act, and adapt with minimal human input. It's a testbed for ideas on reliability, tool orchestration, and long-running task execution — built to work, not just to demo.
Home Automation
A full-cover smart home built on Home Assistant, tying together lighting, climate, security, and sensors into one coherent, locally-controlled system. It's a real-world deployment target for the AI agent work above — automation that's actually lived in daily.