RAI

RAI is a flexible AI agent framework to develop and deploy Embodied AI features for your robots.

---

---

RAI framework

• rai core: Core functionality for multi-agent system, human-robot interaction and multi-modalities.
• rai whoami: Tool to extract and synthesize robot embodiment information from a structured directory of documentation, images, and URDFs.
• rai_asr: Speech-to-text models and tools.
• rai_tts: Text-to-speech models and tools.
• rai_sim: Package for connecting RAI to simulation environments.
• rai_bench: Benchmarking suite for RAI. Test agents, models, tools, simulators, etc.
• rai_perception: Object detection tools based on open-set models and machine learning techniques.
• rai_nomad: Integration with NoMaD for navigation.
• rai_finetune: Finetune LLMs on your embodied data.