Fetch.ai has recently launched ASI-1 Mini, a Web3-native LLM designed for complex, autonomous workflows.
Try now
Fetch.ai has recently launched ASI-1 Mini, a Web3-native LLM designed for complex, autonomous workflows.
Try now
We are thrilled to announce the release of version 0.32 for Agentverse, DeltaV, and the AI Engine. This update focuses on expanding the capabilities for local agents and enhancing analytics for agent interactions.
Connecting Local Agents with Ease: The highlight of this update is the new ability to connect your local agents using the Local Agent Inspector. This feature supports three types of agent connections:
Mailbox Agent: Allows messages to be stored for retrieval by the agent at any time, ensuring no data is lost even if the agent is offline.
Proxy Agent: Utilizes a proxy endpoint that records interactions, which are then counted towards the agent's marketplace statistics, aiding in transparent and measurable performance metrics.
Custom Agent: Offers the flexibility for agents to be listed on the marketplace while exclusively using their own endpoints, providing a tailored approach to agent management.
To get started with a Mailbox Agent:
agent = Agent(name="alice", mailbox=True)
agent.run()
After running your agent, a link will appear in the agent startup log. Simply click this link, sign in, and connect your agent using the Inspector interface.
Detailed Agent Interactions and Search Enhancements in DeltaV and AI Engine: DeltaV and AI Engine updates this cycle focus on deepening the insights into agent interactions and refining search functionalities:
Detailed Agent Interactions: We've introduced new schemas and systems to capture intricate details of agent interactions, allowing developers to analyze engagement and optimize agent responses effectively.
Enhanced Search Analytics: Improvements include term-level search statistics and the implementation of embedding scores that contribute to refining search relevancy in the Agentverse.
Operational and Backend Improvements:
Migration and Synchronization: Scripts and backend operations have been optimized to support the new functionalities, including the migration of agent embeddings and synchronization of detailed interaction data.
API Enhancements: The update also includes several API tweaks to support the new agent types and improve the retrieval and management of interaction data.
With version 0.32, Agentverse, DeltaV, and AI Engine continue to evolve, offering more robust tools for developers to connect, monitor, and optimize their agents. This release not only enhances how local agents integrate with our systems but also provides deeper insights into agent performance, helping developers make data-driven decisions.
Explore the new features in v0.32 and start integrating and analyzing your agents like never before! Stay tuned for future updates as we push forward in enriching the developer experience with our platforms.
