This is an open-sourced, research-driven initiative designed to explore and develop multi-agent AI systems over the winter. The project focuses on building a General Purpose AI Agent Framework and establishing a Registry of Domain-Specific AI Agents.
- Intros
- AI Agent Project Overview
- Agent Presentation
- Agent Registry
- Agent Product Specs
- Codebase Setup
- Communications / Next Steps
- Q&A
- Framework: General Purpose AI Agent Framework
- Registry: A collection of specialized AI agents tailored to specific domains.
- Comprehensive guides for understanding and using the framework and registry.
- A user-friendly interface for showcasing the project.
- Frontend/UI: Intuitive design for interaction.
- Backend: Scalable and efficient architecture.
- Software Developer Kit (SDK): Tools for developers to build and integrate agents.
- A structured catalog of reusable, domain-specific AI agents.
- Contribute to the open-source agent repository or develop a closed-source agent to add to the registry.
- Deliver a Final Presentation during “AI Day,” showcasing:
- Video tutorials of AI agents in action.
- Research insights from the project.
The project enables various ways to deploy and use AI agent systems:
- Access through the product UI.
- Deploy via a desktop application.
- Use the developer kit to integrate into applications.
- NLP queries are routed to one or more agents (Agent 1, Agent 2, Agent 3, etc.).
- Tasks are performed either in series or in parallel.
- The system generates and returns an output to the user.
- Includes logs to monitor agent progress.
- Spawn agent templates directly into the chatbot.
- Use a Drag & Drop Flow to design and execute compound agent systems.
Join the initiative by contributing to the open-source repository, building specialized agents, or collaborating on research and development. Showcase your work during the AI Day event!
Note: to run code correctly you'll need to replace the openAI api key