Twin Robo HomeMind demonstrates a coordinated AI household system where a planning agent generates structured task sequences using an open-weight LLM, and an execution agent performs and reports task lifecycle events via a live dashboard.
Python modular architecture, GPT-OSS-20B via Together.ai, file-based JSON message bus, Streamlit dashboard with human-in-the-loop approval controls.
AI caregiving systems, assistive robotics workflows, smart home task automation, human-in-the-loop AI supervision.
To expand into real-world robotics using MQTT/ROS communication, live perception models (YOLO/OCR), and IoT integration.
Python, Streamlit, Together.ai API, JSON bus, pandas, requests