Agentic 2
satya - 2/22/2025, 2:29:28 PM
Summary
- pydantic-ai
- huggingface smolagents
- storm
- autogen
- AG2 (suggested as a successor to AutoGen)
- CREO-1
- Atomic Agents (by BrainBlend-AI)
satya - 2/22/2025, 3:29:05 PM
Langroid: https://github.com/langroid/langroid
satya - 2/22/2025, 3:33:27 PM
Comparing AG, LangGraph, and Crew
satya - 2/22/2025, 3:36:02 PM
Key adoption in the order of the 3
- AG (30k)
- Crew (20k)
- LangGraph (10k)
satya - 2/22/2025, 3:36:29 PM
Features
- Streaming
- Human loop
- Time travel to look back
- Memory
- Low code
satya - 2/22/2025, 3:37:50 PM
AutoGen architecture diagrams
satya - 2/22/2025, 3:41:29 PM
Briefly
- Actor based
- Message passing
satya - 2/22/2025, 3:42:55 PM
Also based on chats
Also based on chats
satya - 2/22/2025, 3:43:05 PM
AutoGen getting started
AutoGen getting started
satya - 2/22/2025, 3:45:04 PM
An example of an agent convesations
satya - 2/22/2025, 3:45:39 PM
AG Learning curve: 10/10, fast
AG Learning curve: 10/10, fast
satya - 2/22/2025, 6:45:32 PM
MS Autogen Feature list
MS Autogen Feature list
satya - 2/22/2025, 7:25:39 PM
New focus
- complete redesign of the AutoGen library, aimed at improving
- code quality,
- robustness,
- generality,
- and the scalability of agentic workflows.
satya - 2/22/2025, 7:26:04 PM
Features are listed here on MS site
satya - 2/22/2025, 7:26:26 PM
Layers
satya - 2/22/2025, 7:27:27 PM
Feaures
- Asynchronous messaging: Agents communicate through asynchronous messages, supporting both event-driven and request/response interaction patterns.
- Modular and extensible: Users can easily customize systems with pluggable components, including custom agents, tools, memory, and models. They can also build proactive and long-running agents.
- Observability and debugging: Built-in tools provide tracking, tracing, and debugging agent interactions and workflows, with support for OpenTelemetry for industry-standard observability.
- Scalable and distributed: Users can design complex, distributed agent networks that operate seamlessly across organizational boundaries.
- Built-in and community extensions: The extensions module enhances the framework?s functionality with advanced model clients, agents, multi-agent teams, and tools for agentic workflows. Support for community extensions allows open-source developers to manage their own extensions.
- Cross-language support: This update enables interoperability between agents built in different programming languages, with current support for Python and .NET and additional languages in development.
- Full type support: Interfaces now enforce type checks at build time, ensuring robust and cohesive code quality.