Skip to main content

AI Agent Studio Document Guidelines: To Maximize Agent Accuracy

Document Guidelines

The content and structure of a document are key factors in retrieving accurate answers to queries.



1. Document Content

Composition

  • Each document is a logical collection of attachments (files).

Supported Formats

  • PDF (must meet accessibility standards)

  • Text (.txt)

Language

  • Only English is supported.

Size & Limits

  • Each attachment: up to 300 pages

  • Each document: up to 500 attachments

  • Total size across all attachments: ≤ 2 GB

Limitations

  • Multi-modal questions not supported:

    • Only text-based content will be answered.

    • Questions about pictures, graphics, charts, etc. will not be answered.

  • Complex tables spanning multiple pages are not supported.



2. Document Structure

Importance

A clear structure allows the agent to split the document into semantically meaningful chunks, improving the quality of answers.

Best Practices

  • Use headings and subheadings to organize content.

  • Headings must clearly describe the meaning of their section.

  • Each section should contain related concepts only.

    • If multiple concepts are mixed, split into new sections.

  • Ensure heading levels are visually distinguishable.

Example of Heading Levels

  • Heading 1 → Font size 12

  • Heading 2 → Font size 10

  • Heading 3 → Font size 8

  • (and so on…)

Comments

Post a Comment

Popular posts from this blog

Oracle Fusion AI Agent Studio 25D: Pre-Built AI Agents across CX, GRC, HCM, PROC, PSC and SCM Cloud.

  Oracle has delivered an impressive set of 124 Pre-Built AI Agents across CX, HCM, Procurement, SCM, GRC, and PSC as of release 25D. These agents are designed to accelerate productivity, reduce manual effort, and bring intelligent automation directly into day-to-day business processes. Organizations can now tap into ready-to-use AI capabilities without heavy development or custom modeling. Each agent is crafted to solve a specific business scenario with Oracle’s embedded AI and contextual intelligence. ✅ CX users benefit from smarter case handling and service automation. ✅ HCM teams gain intelligent assistance in recruiting, payroll, and workforce tasks. ✅ Procurement and SCM see boosted efficiency with automated validations, insights, and exception handling. ✅ GRC and PSC get enhanced compliance, risk detection, and operational governance. These pre-built agents also serve as a foundation for extending or developing custom agents in Oracle AI Agent Studio. 25D marks a ma...
Post a Comment
Read more

Deploying an AI Agent in Oracle Fusion Applications Page

  πŸš€ 𝐸π‘₯𝑐𝑖𝑑𝑒𝑑 π‘‘π‘œ π‘ β„Žπ‘Žπ‘Ÿπ‘’ π’‰π’π’˜ 𝑢𝒓𝒂𝒄𝒍𝒆 𝑨𝑰 π‘¨π’ˆπ’†π’π’• π‘Ίπ’•π’–π’…π’Šπ’ π’Šπ’” π’•π’“π’‚π’π’”π’‡π’π’“π’Žπ’Šπ’π’ˆ π‘‘β„Žπ‘’ π‘€π‘Žπ‘¦ 𝑀𝑒 π‘€π‘œπ‘Ÿπ‘˜! πŸ€– π‘«π’†π’‘π’π’π’šπ’Šπ’π’ˆ 𝑨𝑰 π‘¨π’ˆπ’†π’π’•π’” π’…π’Šπ’“π’†π’„π’•π’π’š π’Šπ’π’•π’ π‘­π’–π’”π’Šπ’π’ π‘·π’‚π’ˆπ’†π’” π‘’π‘›π‘™π‘œπ‘π‘˜π‘  π‘ π‘šπ‘Žπ‘Ÿπ‘‘π‘’π‘Ÿ π‘€π‘œπ‘Ÿπ‘˜π‘“π‘™π‘œπ‘€π‘  π‘Žπ‘›π‘‘ π‘“π‘Žπ‘ π‘‘π‘’π‘Ÿ π‘‘π‘’π‘π‘–π‘ π‘–π‘œπ‘›π‘ . ✨ π‘‡β„Žπ‘’ π‘“π‘’π‘‘π‘’π‘Ÿπ‘’ π‘œπ‘“ 𝑖𝑛𝑑𝑒𝑙𝑙𝑖𝑔𝑒𝑛𝑑 π‘’π‘›π‘‘π‘’π‘Ÿπ‘π‘Ÿπ‘–π‘ π‘’ π‘Žπ‘’π‘‘π‘œπ‘šπ‘Žπ‘‘π‘–π‘œπ‘› 𝑖𝑠 β„Žπ‘’π‘Ÿπ‘’! Note: Not all pages are available for now. Mostly it is CX Cloud pages available for this.    
Post a Comment
Read more

AI Agents and Agentic AI: Redefining the Oracle Enterprise Applications Landscape

  Artificial Intelligence is no longer a futuristic concept—it is actively reshaping how organizations operate . In the domain of Oracle Enterprise Applications—including ERP, SCM, HCM, CX, and EPM —AI is evolving from a supportive tool into a strategic enabler . Among the most transformative developments are AI Agents and Agentic AI , which are set to redefine how users interact with enterprise systems, make decisions, and drive outcomes . Understanding AI Agents and Agentic AI AI Agents are task-specific assistants designed to streamline and optimize routine processes. They can automate repetitive tasks, provide actionable insights, and support decision-making. For example, an AI Agent might suggest invoice approvals in ERP, forecast inventory in SCM, or recommend training programs in HCM . Agentic AI represents the next frontier of intelligence— autonomous systems capable of reasoning, planning, and executing multi-step workflows across applications . Unlike task-focus...
Post a Comment
Read more