Popular AI orchestration framework LlamaIndex introduces 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 evolve, this approach may offer organizations an option for improving agents’ decision-making capabilities.

LlamaIndex says ADW helps agents manage “complex workflows that go beyond simple acquisition or matching.”

Some agent frameworks are based on RAG systems, which provide agents with the information they need to complete tasks. However, this approach does not allow agents to make decisions based on this information.

LlamaIndex provides some real-world examples of how ADW works well. For example, in contract reviews, human analysts must extract important information, cross-reference regulatory requirements, identify potential risks and generate recommendations. When deployed in that workflow, AI agents better follow the same pattern and make decisions based on the documents they read for contract review and knowledge from other documents.

“ADW addresses these challenges by treating documents as part of a broader business process,” LlamaIndex said in a blog post. “An ADW system can maintain the state of steps, apply business rules, coordinate different components and take actions based on document content – not just analyze it.”

LlamaIndex previously stated that RAG, while an important technique, remains primitiveespecially for businesses looking for stronger decision-making capabilities using AI.

Understanding the context for decision making

LlamaIndex has developed a reference architecture that integrates LlamaCloud’s agent parsing capabilities. It “builds systems that understand context, maintain state and drive multi-step processes.”

To do this, each workflow has a document that acts as an orchestra. It can direct agents to tap into LlamaParse to extract information from the data, maintain the context and process status of the document, then retrieve reference material from another knowledge base. From here, agents can begin making recommendations for use case contract review 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 acquisition or matching,” the company said. “This approach allows them to create deep context about the documents they are processing while coordinating between different system components.”

Different agent frameworks

Agent orchestra is an evolving space, and many organizations are still exploring how agents – or multiple agents – work for them. Orchestrating AI agents and applications becomes possible bigger conversation this year as agents move from a single system to a multi-agent ecosystem.

AI agents are an extension of what RAG offers, namely, the ability to search for information based on business intelligence.

But as more businesses begin to deploy AI agents, they also want them to perform many of the tasks performed by human employees. And, for these more complex use cases, “vanilla” RAG is not enough. One of the advanced methods that businesses are considering is agent RAGwhich expands the agents’ knowledge base. Models can decide if they need to find more information, which tool to use to get that information and if the context it’s taken from is relevant, before arriving at a result.

Daily insights into business use cases in VB Daily

If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory changes to practical deployments, so you can share insights for maximum ROI.

Read our Privacy Policy

Thanks for subscribing. See more VB newsletters here.

An error occurred.