Agents Store

Array ( [0] => Array ( [_id] => 677e3112a901830024291f54 [name] => Dynamic Knowledge Base Creation Agent [description] =>

ZBrain Dynamic Knowledge Base Creation Agent automates the maintenance and continuous updating of organizational knowledge bases. Leveraging a Large Language Model (LLM) and advanced technologies, the agent ensures that knowledge repositories are always current by validating URLs, detecting content changes, and maintaining an up-to-date knowledge base.

Challenges the Dynamic Knowledge Base Creation Agent Addresses:

The rapid evolution of information and the labor-intensive demands of manual updates often hamper most organizations' efforts to keep their knowledge bases accurate and current. This often leads to the dissemination of outdated or incorrect information, increased workload for staff managing content updates, delays in critical decision-making, and inconsistency across departmental information systems. Such challenges undermine efficiency, reduce productivity, and frustrate both employees and customers.

ZBrain Dynamic Knowledge Base Creation Agent transforms knowledge management by leveraging an LLM and advanced technologies to monitor, identify, and assimilate new data into existing knowledge bases without human intervention. By automating these processes, the agent eliminates manual errors, reduces team workload, and ensures that all stakeholders have access to the most current and accurate information. This not only improves decision-making and customer support but also fosters a more agile and responsive organizational structure.

How the Agent Works?

The agent follows a structured, step-by-step process to ensure accuracy, prevent redundancy, and streamline knowledge management. Below is a detailed breakdown of how the agent processes documents.

Step 1: User Input & URL Capture

The process begins when a user submits a list of URLs that point to documents intended for addition or update in the knowledge base. These documents can include guidelines, policies, contracts, reports, or other essential digital files.

Key Tasks:

Accepting Multiple URLs: Users can submit one or more URLs at a time, allowing for batch processing.
Direct Integration: URLs can be manually entered via the agent interface.
Queue Management: Submitted URLs are temporarily stored in a queue for processing.
Real-time Trigger: The agent automatically initiates validation and processing upon URL submission, eliminating manual intervention.

Outcome:

The agent successfully collects URLs and prepares them for validation.

Step 2: URL Validation & Processing

Once URLs are received, the agent validates them for correctness, accessibility, and relevance.

Key Tasks:

Format Verification: The agent checks if each URL is properly structured and accessible.
LLM-based Formatting: The agent structures all URLs into a loop-friendly format for efficient processing.
Error Handling: If a URL is invalid, non-existent, or irrelevant, the agent flags it as an error and stops further processing for that entry.
Proceeding with Valid URLs: Only verified URLs move forward to the next stage.

Outcome:

The agent filters out invalid URLs, ensuring only relevant documents are processed.

Step 3: Knowledge Base Cross-check

The agent scans the KB to check if a document corresponding to the submitted URL already exists.

Key Tasks:

Searching for Existing Entries: The agent queries the knowledge base using the submitted URL to determine if a matching document is already stored.
Retrieving the Knowledge Base ID: If a match is found, the agent retrieves the document’s unique ID for further analysis.
Assigning a New ID: If no match is found, the agent generates and assigns a new unique ID to the document before adding it to the knowledge base.

Outcome:

The agent prevents duplicate uploads and ensures that only new or updated files are stored.

Step 4: Intelligent Content Matching

For URLs linked to existing documents, the agent performs a hash comparison to determine whether the content has changed.

Key Tasks:

Generating Unique Hashes: The agent calculates a hash for both the existing document and the newly submitted file.
Comparing Hash Values:
- If the Hashes are Different -The document has been modified, so the agent replaces the old version with the updated file.
- If the Hashes are Identical - The document remains unchanged, and no further action is required.
Handling New Documents: If a document does not exist in the KB, the agent automatically uploads it as a new entry.

Outcome:

The agent prevents unnecessary uploads while ensuring only updated files are stored.

Step 5: Content Analysis

To confirm and summarize changes, the agent leverages a Large Language Model (LLM) for content comparison.

Key Tasks:

The agent receives two versions of the document for analysis:
- Old Content – The existing version stored in the knowledge base.
- New Content – The updated document submitted via URL.
Processing Steps:
- Comparison & Analysis:
  - Performs a comprehensive comparison to detect modifications, additions, and deletions.
  - Examines changes at a granular level to assess the extent of updates.
- Change Detection & Categorization:
  - Identifies key differences between the old and new versions.
  - Classifies modifications based on content structure, such as text updates, formatting changes, or metadata adjustments.
- Structured Summary Generation:
  - Overview of Changes – Provides a high-level summary outlining key differences.
  - Detailed Modifications – Highlights specific sections, lines, or content that have been added, modified, or removed.

Outcome:

The agent provides a clear and concise summary of document updates, enabling users to track modifications efficiently.

Why use the Dynamic knowledge base creation agent?

Automated Updates: Automatically monitors and updates documents in the knowledge base.
Reduced Manual Effort: Eliminates the need for manual document uploads and version checks.
Efficient Document Comparison: Uses hashing to detect document changes and prevent duplication.
Real-Time Processing: Ensures the knowledge base is updated in real time.
Seamless Integration: Direct integration with the KB for smooth document management.
Scalable Workflow: Handles large volumes of documents and URLs with ease.
Version Control: Automatically replaces outdated documents with the latest versions.

[image] => https://d3tfuasmf2hsy5.cloudfront.net/assets/worker-templates/technician-assignment-agent.svg [icon] => https://d3tfuasmf2hsy5.cloudfront.net/assets/worker-templates/technician-assignment-agent.svg [sourceType] => FILE [status] => READY [department] => Utilities [subDepartment] => Dynamic Knowledge Creation [process] => Knowledge Base Update [subtitle] => Creates and updates a knowledge base based on provided input resources, ensuring that the information remains current and comprehensive. [route] => dynamic-knowledge-base-creation-agent [addedOn] => 1736323346984 [modifiedOn] => 1736323346984 ) [1] => Array ( [_id] => 6814a5eb684a1282b8e6965f [name] => Jira Conversational Insights Agent [description] => The Jira Based Conversational Agent enables users to interact with Jira data using natural language, transforming how engineering, operations, and support teams access information. Instead of relying solely on Jira Query Language (JQL) or manual filtering, users can simply ask questions in plain language to retrieve insights from issues, attachments, comments, and linked documentation.

The agent combines advanced natural language processing (NLP), semantic search, and JQL interpretation to understand user intent and return relevant, context-rich results. It processes structured and unstructured data across multiple projects, intelligently surfacing information such as ticket histories, resolution steps, related SOPs, and team discussions—without the need to manually navigate through the Jira interface.

This conversational interface accelerates knowledge discovery and reduces time spent on repetitive searches or escalations. It supports real-time use cases, including incident response, sprint planning, and onboarding, and continuously improves its accuracy through feedback loops and usage patterns. By enabling faster, smarter access to operational insights, the Jira Data Conversational Query Agent empowers teams to make informed decisions and scale knowledge sharing across the organization.

[image] => https://d3tfuasmf2hsy5.cloudfront.net/assets/worker-templates/lead-qualification-scoring-worker.svg [icon] => https://d3tfuasmf2hsy5.cloudfront.net/assets/worker-templates/lead-qualification-scoring-worker.svg [sourceType] => FILE [status] => REQUEST [department] => Utilities [subDepartment] => Dynamic Knowledge Creation [process] => Knowledge Base Management [subtitle] => Leverages JQL and NLP to provide quick, context-driven insights from Jira tickets, attachments, and procedural documents. [route] => jira-conversational-insights-agent [addedOn] => 1746183659470 [modifiedOn] => 1746183659470 ) [2] => Array ( [_id] => 67cec35d2e7f0a02273d3289 [name] => Salesforce Knowledge Creation Agent [description] =>

The Salesforce Knowledge Creation Agent automates the process of generating and managing knowledge base articles from existing case data. It streamlines the conversion of complex case data into easily accessible knowledge resources, ensuring valuable troubleshooting information is consistently captured, accurately formatted, and efficiently stored within the knowledge base. This enhances customer support effectiveness and empowers self-service capabilities, making information retrieval quicker and more reliable for support teams.

Challenges the Agent Addresses

Manually creating and maintaining knowledge articles can be both time-consuming and prone to errors, especially in fast-paced environments where a high volume of customer service cases is processed daily. Without an automated system, important case details may not be captured effectively, leading to missed opportunities for valuable insights that could aid future issue resolution. Additionally, the risk of duplicate articles cluttering the knowledge base makes it harder for customer agents to find relevant information quickly.

The Salesforce Knowledge Creation Agent addresses these challenges by automatically generating well-structured knowledge articles, ensuring that sensitive customer information is redacted, and preventing duplicate entries, streamlining the entire process for improved efficiency and accuracy.

How the Agent Works

The Salesforce Knowledge Creation Agent automates and optimizes the process of generating knowledge articles, ensuring high standards of consistency, accuracy, and efficiency. The agent is triggered whenever a new request for knowledge content is submitted or when incoming cases are received. Leveraging an LLM, the agent intelligently analyzes incoming data, creates relevant and well-structured articles, and ensures seamless integration with Salesforce's knowledge management standards. Below is a detailed breakdown of how the agent functions:

Step 1: Case Data Retrieval and Processing

The process begins when a case is received through an integrated system. The agent fetches all relevant case details and prepares them for further processing.

Key Tasks:

Case Data Extraction: The agent retrieves case information, including the case number, description, and other contextual details.
Data Structuring: The extracted case data in JSON format is transformed into a standardized, ServiceNow-compatible structure using an LLM for seamless processing.

Outcome:

The agent successfully gathers and structures case data, ensuring it is ready for the next steps.

Step 2: PII Guardrails and Data Redaction

To ensure compliance and protect customer privacy, the agent applies PII (Personally Identifiable Information) guardrails to remove sensitive details from the case data.

Key Tasks:

Detection of Sensitive Information: The agent identifies PII such as customer names, phone numbers, email addresses, and account numbers from case details using an LLM.
Automated Redaction: Any detected PII is removed before proceeding.
Validation Check: The agent ensures that only non-sensitive, relevant case details remain for the knowledge article.

Outcome:

The processed case data is free of sensitive customer information and ready for knowledge article generation.

Step 3: Knowledge Article Formatting

The agent converts the structured case data into a knowledge article format.

Key Tasks:

Markdown Structuring: The agent organizes case information into a clear, standardized format for improved readability and consistency.
HTML Conversion: The Markdown-formatted data is converted into HTML for seamless integration with the knowledge base system.

Outcome:

The case details are structured and formatted for easy comprehension.

Step 4: Duplicate Knowledge Article Check

Before creating a new knowledge article, the agent checks whether an article already exists for the given case to prevent duplication.

Key Tasks:

Fetching Existing Articles: The agent retrieves a list of all existing knowledge articles from the knowledge base.
Title Matching: The agent compares the titles of existing articles with the current case title and case ID to check for duplicates.
Duplicate Verification: If an article with the same case ID already exists, the agent flags it as a duplicate.

Outcome:

If an existing article is found, the agent retrieves and provides the existing article’s URL.
If no existing article is found, the agent proceeds to create a new one.

Step 5: Knowledge Article Creation and Publishing

If no duplicate article exists, the agent proceeds to create and publish a new knowledge article.

Key Tasks:

API Call to Knowledge Management System: The agent sends a request to the ServiceNow API to create a new article.
Content Submission: The agent submits the formatted case details.
Confirmation and URL Generation: Once created, the system generates a unique URL for the knowledge article.

Outcome:

A new knowledge article is successfully created and published.
The generated URL is returned for future reference, improving efficiency and accessibility.

Why Use the Salesforce Knowledge Creation Agent?

Automates Article Creation: Reduces manual effort by generating structured knowledge articles, allowing support teams to focus on resolving new cases.
Enhances Knowledge Base Accuracy: Publishes only verified, well-structured, and duplicate-free content to maintain high-quality documentation.
Faster Resolution and Response Times: Provides instant access to relevant knowledge articles, helping agents resolve similar cases quickly and improving overall service response times.
Ensures Compliance and Data Privacy: Applies robust PII detection and redaction to safeguard sensitive customer information.
Seamless Salesforce Integration: Works natively within Salesforce, enabling real-time knowledge management without disrupting workflows.
Scalable and Customizable: Adapts to various case types and business needs, allowing for tailored workflows and flexible knowledge article formats.