A standard, open framework for building AI agents is coming from Cisco, LangChain and Galileo

One of the goals of the agent’s future is for artificial intelligence agents from various organizations to speaking freely and smoothly with each other. But reaching this point requires the susceptibility of inter -operating, and these agents may have been designed with LLMS, data frameworks and different symbols.

To achieve interconnection, the developers of these agents must agree on how they communicate with each other. This is a difficult task.

A group of companies, including Cisco, Langchain, Llamaindex, Galileo and Glean, created Agntcy, an open source group with the aim of creating a possibility to operate between the agent. Agntcy aims to facilitate any agent of Amnesty International to communicate and exchange data with another.

Unification of artificial intelligence agents

“Just as when the cloud, the Internet, the acceleration of applications and all social interactions on a global scale, we want to build the Internet of agents who speed up all human work on a global scale,” said Vigoy Bandy, president of Cisco, in an interview with Vesorebeat.

The AGNTCY likened to the appearance of the TCP/IP/IP control protocol (TCP/IP) and the DNS name system (DNS), which helped regulate the Internet and allow interdependence between different computer systems.

He said: “The way we think about this problem is that the original internet allowed humans, servers and web farms with everyone.” “This is the Internet of agents, and the only way to do this is to make it open and interim operation.”

Cisco, Langchain and Galileo will work as a basic governorate for Agntcy, with Glean and Llamaindex as shareholders. However, this structure may change as the collective adds more members.

Unification of a fast -moving industry

Artificial intelligence agents cannot be a island. To reach their full potential, they should be able to communicate with other agents who lie outside the institution’s network. This is where the inter -operation comes.

Putting standards in traditional industries is difficult enough; It becomes more difficult for technology like artificial intelligence, as typical promotions and changes occur every few months. However, this is not the first time that a standard for obstetric males has been proposed.

In Langchain, one of the primary Agntcy members, has its own protocol to work with agents built on frameworks other than Langchain. The agent protocol, which was launched in November last year, allows Langchain to speak to the agents who were created using Autogen, Crewai or any other framework.

Meanwhile, Anthropor announced the typical context protocol (MCP) in November. This protocol aims to unify how artificial models and tools connect to data sources. But while many developers adopted MCP, it is not a complete standard yet.

Yash Seth, founder of the artificial intelligence evaluation platform Galileo, said the measure is “very important.”

“In fact, unification is needed, it will increase the speed of the adoption of agents. Today, the difference is built in silos, you should discover how to develop their infrastructure components from zero point,” Seth said in an email. “The unification of multiple mold systems can only occur if these factors that work with non -live models have a strong anchor in measuring and informing their performance, accuracy and reliability.”

Seth admits that making artificial intelligence customers can be complicated. Agntcy “wants to encourage developers to expand these specifications and applications programming and tools to suit their needs instead of re -invention of the wheel, which will be decisive to achieve monotheism.”

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

“Building agents are already possible, and is done.

A platform and language simultaneously

Agntcy is more than just a set of codes for agents. Customers will also allow the discovery of agents from different developers who manage the Agntcy standard.

“Customers can sew all of these agents on the Agntcy platform so that they can discover, publish and evaluate while building their workflow tasks,” Pandnesse said.

Agntcy still needs to employ more artificial intelligence players to add new agents to the platform and gain momentum as a standard. After all, in order for something to become a standard in industry, there must be a collective dependence, to prevent the creation of many competing criteria.

This is where you face projects like Agntcy, an arduous battle. Pandnesse said that the group is talking to many other players in the industry, and they want to get the largest possible number of views while developing the platform. It will take some time.

Meanwhile, institutions continue to experience and even publish artificial intelligence agents. Perhaps in the future, everything will be able to talk to each other.