Join our daily and weekly newsletters for the latest updates and exclusive content covering cutting-edge AI. Learn more
Popular AI orchestration framework LamaIndex introduced Agent Document Workflow (ADW), a new architecture that the company says goes beyond retrieval augmented generation (RAG) processes and increases agent productivity.
As orchestration frameworks continue to improve, this method could provide organizations with an option to improve agents’ decision-making capabilities.
LlamaIndex claims that ADW can help agents manage “complex workflows beyond simple extraction or matching.”
Some agentic frameworks are based on RAG systems, which provide agents with the information they need to accomplish their tasks. However, this method does not allow agents to make decisions based on this information.
LlamaIndex gave some concrete examples of how ADW would work well. For example, during contract reviews, human analysts must extract key information, cross-reference regulatory requirements, identify potential risks and generate recommendations. When deployed in this workflow, AI agents would ideally follow the same model and make decisions based on the documents they read to review the contract and their knowledge gleaned from other documents.
“ADW addresses these challenges by processing documents as part of broader business processes,” LlamaIndex said in a statement. blog post. “An ADW system can maintain state across stages, apply business rules, coordinate different components, and take actions based on document content, not just parse it. »
LlamaIndex has previously stated that RAG, while an important technique, remains primitive, especially for companies seeking more robust decision-making capabilities through AI.
Understanding context for decision making
LlamaIndex has developed reference architectures combining its LlamaCloud analytics capabilities with agents. It “builds systems that can understand context, maintain state, and drive multi-step processes.”
To do this, each workflow has a document which acts as an orchestrator. It can instruct agents to leverage LlamaParse to extract information from data, maintain document context and process state, and then retrieve reference material from another knowledge base. From there, agents can start generating recommendations for the contract review use case or other actionable decisions for different use cases.
“By maintaining state throughout the process, agents can manage complex, multi-step workflows that go beyond simple extraction or matching,” the company said. “This approach allows them to create deep context on the documents they process while ensuring coordination between the different components of the system. »
Different agent frameworks
Agentic orchestration is an emerging space, and many organizations are still exploring how agents (or multiple agents) work for them. Orchestration of agents and AI applications could become a bigger topic of discussion this year, as agents move from single systems to multi-agent ecosystems.
AI agents are an extension of what RAG offers, which is the ability to find insights based on business knowledge.
But as more companies begin to deploy AI agents, they also want them to perform many of the tasks done by human employees. And, for these more complicated use cases, the “vanilla” RAG is not enough. One of the advanced approaches companies are considering is agentic RAG, which expands the knowledge base of agents. Models can decide whether they need to find more information, which tool to use to obtain that information, and whether the context they just extracted is relevant, before arriving at a result.
