A virtual assistant that makes your life better! ¶
-
Aeiva augments human intelligence
We focus on developing artificial intelligence techniques that augment human intelligence rather than replace human.
-
Multimodal and embodied agent
Aeiva aims to support multimodal and embodied environments. Currently, it can process text, image, audio, and video inputs.
-
AI for science
Aeiva aims to be an expert agent that equipped with professional skills and can help accelerate scientific discovery. We currently focus on materials science and health.
-
Safe, controllable, and interpretable
We emphasize on the safety, controllability, and interpretability (SCI) of AI. We aim to incorporate SCI techniques in Aeiva, as well as developing and evaluating SCI techniques based on it.
-
AI community and society
We aim to construct multi-agent AI community and society for self-evolving AI.
-
Open source
We open source our code on GitHub. Licensed under Apache license 2.0.
Examples of what Aeiva can do ¶
Multimodal Chatbot
Aeiva can process multimodal inputs and chat with you. To be updated ...
Intelligent Desktop Assistant
Aeiva can serve as an work assistant on your desktop. You can interact with it by voice or textual inputs. It supports multilingual TTS and STT, as well as basic computer use functionalities. The functionality will be keep improved. Stay tuned ...
Multi-Agent Community
Aeiva aims to support multi-agent AI community research. To be updated ...
Learn more
AI for Science 
The ultimate goal of Aeiva is supporting the research of AI and accelerate scientific research and discovery. To be updated ...
Learn more
