Delivered AI Agents
The delivered AI agents provide predefined AI functionality using Oracle AI for Fusion Applications.
Your agency can use delivered agents out-of-the-box or as templates to create your own. For more information about setup, including embedding the AI Assist button for access to agents, see Integrating with AI Agents. Some agents require the agency to provide a Retrieval-Augmented Generation (RAG) document, which enhances a large language model's (LLM) responses by providing it with external, factual information.
Let's take a look at the delivered AI agents for Oracle Permitting and Licensing:
|
AI Agent |
Product |
Pages |
Description |
|---|---|---|---|
|
Agency Business License Assistant |
Business Licenses |
Requires agency setup on extensible pages: Licenses list page and license details |
Retrieves business license application-related information. |
|
Agency Permit Assistant |
Permits |
Requires agency setup on extensible pages: Permit Applications list page and permit details |
Uses the Permit Application Finder tool to retrieve permits and permit-related information. |
|
Agency Planning Assistant |
Planning and Zoning |
Requires agency setup on extensible pages: Planning Applications list page and planning application details |
Retrieves planning application-related information. |
|
Assign Planner |
Planning and Zoning |
Requires agency setup on extensible pages: Planning Applications list page |
Helps agency staff find the best planner for a planning and zoning application. |
|
Code Enforcement Case Assistant |
Code Enforcement |
Requires agency setup on extensible pages |
Helps agency staff to do quick inquiries on cases like unassigned cases, active cases, all cases, and case time lines. |
|
Code Enforcement Reported Incident Checker |
Code Enforcement |
Requires agency setup on extensible pages |
Leverages an agentic workflow to automate the incident review process, including identifying duplicate incidents, assigning cases based on workload, and sending assignment emails. |
|
Code Incident Analyzer |
Code Enforcement |
Requires agency setup on extensible pages |
Analyzes incidents to provide a summary to agency staff based on questions asked about incoming, overdue, and unassigned incidents. |
|
Contractor License Assistant |
Permits |
Requires agency setup on extensible pages: Contractors list page and contractor license details |
Retrieves contractor licenses and related information, including contractor license facts. |
|
Inspection Summary |
Common Components |
Requires agency setup on extensible pages. |
Summarize an inspection and output in formatted HTML. |
|
Permit Inspection Assistant |
Permits |
Requires agency setup on extensible pages: Pages in the Inspector Application, including Inspection Tasks, Inspection Overview, Inspection Result |
Retrieves inspections, inspection comments, inspection checklists, inspection checklist items, and instructions. The agent has access to inspections, inspection comments, checklist and checklist items, and business objects to provide users with inspection details. |
|
Planning Activity Assistant |
Planning and Zoning |
Requires agency setup on extensible pages. |
Retrieves planning activities associated with a planning and zoning application. |
|
Plan Review PreScreener |
Common Components |
Requires action tile setup on registered public user landing page for Permits and Planning and Zoning |
Features AI image recognition and code comparison to assess the completeness and currency of a submitted plan set. |
|
Preapplication Agent |
Planning and Zoning |
Preapplication Outcome Page |
Combines the application data, zoning district information, and requirements from the municipal codes to deliver the preapplication outcome response in natural language to the user. The agency provides the Retrieval-Augmented Generation (RAG) document for this agent. |
|
Public Document Agent |
Common Components |
Registered User Landing Page Guest User Landing Page Apply Page |
Designed to provide precise, reliable answers strictly based on the contents of the provided documents. The assistant references specific sections or excerpts when possible. If an answer can't be found within the provided materials, the assistant clearly states that the information is unavailable. The agency provides the Retrieval-Augmented Generation (RAG) document for this agent. |
|
Reference APO Assistant |
Common Components |
Requires agency setup on extensible pages |
Retrieves APO (address, parcel, and owner) and district information. This agent has access to reference parcels, reference addresses, reference owners, and reference districts tools. |
In the AI Agent Studio, select the PSC family in the search bar to view the available AI agents. Agents in Common Components retrieve application-related information, including permit, business license, and planning and zoning.
Plan Review PreScreener
The Plan Review PreScreener (PLAN_REVIEW_PRESCREENER) agent provides AI-driven analysis featuring image recognition that lets registered public users prescreen plan documents. An action card on the registered user's landing page has been configured with the Plan Review PreScreener agent template so users can upload plan files and receive email results before proceeding with plan review activities.
The Check My Plan Documents card appears under the More actions section on the registered public user's landing page. When selected, the card opens a standard AI prompt drawer over the landing page. From the drawer, users upload one or more plan documents using the paper clip attachment icon. After upload and entering any text in the Ask Oracle prompt field, the AI agent workflow evaluates the documents, displays a success indicator, and generates an email notification with results. The user closes the drawer after entering text or after the success indicator appears; processing can occur in the background.
The Plan Review PreScreener agent template offers AI image recognition and code comparison to assess the completeness and currency of a submitted plan set. Depending on your agency's needs, the agent can inspect common plan package elements such as cover sheet details, professional seals and signatures, owner and project information, code references, site plans, and supporting documentation. It can also cross-reference these items against a standardized preliminary checklist to identify gaps or issues.
Preapplication Agent Team
Oracle delivers the Preapplication Agent Team in the AI Agent Studio and embedded on the Preapplication Outcome page. You can use this template or copy the template to create an agent of your own. The Preapplication Outcome AI Agent helps to analyze the Resource Advisory Council’s (RAC) documents and provide zoning preapplication outcomes. It helps to validate whether the applications and projects meet the local or state government requirements.
For more information about the Preapplication Outcome page, see Working with Preapplication Outcome.