Microservices DAPR now supports artificial intelligence agents

Once again in 2019, Microsoft DAPR open sourceA new operating time to make buildings distributed for small services that are easier. At that time, no one was talking about artificial intelligence agents yet, but as it turned out, DAPR had some basic building blocks to support artificial intelligence agents integrated from the beginning. This is because one of the basic DAPR features is a virtual concept ActorsWhich can receive and process messages, independently of all other actors in the system.

Today, the DAPR team launches DAPR agents, and takes them to help developers to build artificial intelligence agents by providing them with a lot of basic building blocks to do so.

“The agents are a very good use of DAPR,” the co -founder of DAPR and Governor Yaron Schneider explained. “From a technical perspective, you can use the actors as a very light way to operate these agents and be able to operate them on a large scale-and be savings for resources. This is all wonderful, but then, there is still a lot of the logic of work you need to write. There are many frameworks of the agent there, but they do not have the same level of synchronization and the state that DAPR has.

DAPR agents originated from flockA famous open source project, DAPR, for the case of AI’s agent. In an interview with project supervisors, including Microsoft Ai Roberto Rodriguez, the two teams decided to bring the project under the DAPR umbrella to ensure the continuity of the new agent’s framework.

“In many ways, we see the entire agents and terms of the agents about it as another term for” distributed systems, “said the co -founder of DAPR and Governor Mark Fossel.” (…) Instead of summoning accurate services, you can contact them with agents now, mostly because you can put large language models among them all. “

To coordinate these agents efficiently, you need a coincidence engine and a state, as the team argues – which is exactly what DAPR offers. This is partially because DAPR representatives are supposed to be very effective and able to rotate inside a millimeter when a message appears (and closed, while maintaining their condition, when their work is completed).

Currently, DAPR can speak to most famous models outside the box. These include AWS, Openai, Antarbur, the wrong, and embraced face. LLMS support will soon be.

In addition to interacting with these models, given that DAPR agents are the current DAPR frame, developers also get the ability to determine a list of tools that the agent can then use to fulfill a specific task.

Currently, DAPR Python agents support, with. Java, Javascript and GO will be followed soon.