Standard, Open Frame for AI Agents Comes Cisco, Langchain and Galileo

One goal for agent is for AI agents from different organizations to talk freely and seamlessly. But getting to this point requires interoperability and these agents can be built with different LLMs, data frames and code.

In order to achieve operative compatibility, the developers of these agents must agree how they can communicate with each otherS This is a challenging task.

A group of companies including Cisco., Langchain., Llamaindex., Galileo and GleanThey have now set up an AGNTCY, an open source team in order to create a standard standard for agent’s interoperability. Agntcy aims to facilitate any AI agent to communicate and exchange data with another.

Merging agents of AI

“Just as when the cloud and the Internet appeared and accelerated applications and all social interactions on a global scale, we want to build the internet of agents that accelerate all human work globally,” says Vioy Pandy, Head of Cisco, the incubation hand of Cisco,

PANH ARTIFY AGNTCY to the appearance of Transmission/Internet Protocol Control Protocol (TCP/IP) and Domain Name (DNS)which helped to organize the Internet and allow the relationships between different computer systems.

“The way we think about this problem is that the original Internet allows people and servers and web farms to gather,” he said. “This is the internet of agents and the only way to do this is to make it open and operated.”

Cisco, Langchain and Galileo will act as the main supporters of AGNTCY, such as Glean and Llamaindex are associates. However, this structure can change as the team adds more members.

Standardization of a fast moving industry

Ai agents cannot be islands. In order to reach their full potential, they must be able to communicate with other agents that are outside the enterprise network. This is where the interoperability comes here.

Placing standards in traditional industries is sufficiently challenging; It becomes even more difficult for technology like AI, where upgrades and changes in the model happen every few months. However, this is not the first time a standard for generative AI has been proposed.

Langchain, one of the main members of AGNTCY, has its own protocol for working with agents built on frames other than Langchain. Thehe Agent Protocol, Started in November last year, it allows Langchain agents to talk to agents created with autogen, Crewai or other framework.

Meanwhile the anthropic has announced its Model Context (MCP) Protocol in NovemberS This protocol aims to standardize how models and AI tools are connected to data sources. But while many developers have accepted MCP, this is not yet exactly standard.

Yash Sheth, co -founder of the AI ​​Galileo Evaluation Platform, said the standardization “is crucial.”

“Standardization is needed, in fact it will lead to increased speeds to accept agents. Today, teams are being built in silos, and they need to figure out how to develop their own infrastructure components from scratch, “Shet says in an email. “Standardization of multi -agent systems can only happen if these agents powered by non -tender models have a strong anchor when measuring and reporting their performance, accuracy and reliability.”

Shet admits that performing agents for AI operatively compatible can be complicated. Agntcy “wants to encourage developers to expand these specifications, APIs and tools to meet their needs instead of rediscovering the wheel, which will be crucial to achieving standardization.”

Langchain CEO Harrison Chase said in a separate conversation that creating a standard is not impossible, especially now when it is easier to build the agents themselves.

“Construction agents are already possible and are being done. Replit, Klarna, LinkedIn, Uber, Appfolio and many more have already done this. The agents are not the future, they are now. Now that we know how to build agents, the next step is to allow them to connect with each other. This is what the agent’s standard protocol will help to allow, “Chase said.

Platform and language, all at once

Pandey provides AGNTCY as more than just a set of agent codes. This will also allow customers to find agents from different developers who manage the AGNTCY standard.

“Customers can combine all these agents on the Agntcy platform so that they can detect, compose, unfold and evaluate as they build their working streams,” Pandi said.

Agntcy still needs to hire more AI players to add new agents to the platform and get speed as a standard. After all, in order to become an industrial standard, there must be a mass acceptance to prevent too many competitive standards.

It is there that projects like Agntcy are facing a difficult battle. Pandy said the team was talking to many other players in the industry and they want to get as many perspectives as possible while developing the platform. It will take time.

In the meantime, businesses continue to experiment and even deploy AI agents. Maybe in the future they will all be able to talk to each other.