Cassidy AI Essentials

Cassidy AI Essentials

From Knowledge to Action in a day

Cassidy AI gives organizations a practical way to turn internal knowledge, documents, and routine business processes into reliable AI-assisted workflows. Instead of treating artificial intelligence as a separate technical project, the platform allows teams to use familiar information sources to build assistants, automate repetitive tasks, improve consistency, and retrieve critical information more efficiently.

This one-day course concentrates on the capabilities that matter most in real working environments: creating useful agents, connecting approved knowledge, designing controlled workflows, checking the accuracy of generated results, and maintaining proper human oversight. The training will be delivered by an instructor with over 30 years of industry experience and will emphasize current, industry-demanded practices rather than academic theory.

Learning Outcomes

Participants will be able to:

• Navigate the Cassidy AI platform confidently
• Explain the roles of Knowledge Bases, Agents, and Workflows
• Create clear instructions for an AI Agent
• Ground AI responses in approved company information
• Build a basic multi-step workflow using natural language
• Apply Cassidy AI to document review and underwriting support
• Introduce human review and approval controls
• Recognize privacy, accuracy, and governance risks
• Test and improve AI-generated results

Prerequisites

• Basic computer and web-browser skills
• Familiarity with routine business documents and processes
• Access to an approved Cassidy AI workspace
• Sample or non-confidential training documents
• No programming experience required

Training Outline

1. Cassidy AI Fundamentals
   1. Platform purpose and capabilities
   2. Knowledge Bases
   3. AI Agents
   4. AI Workflows
   5. Cassidy Chat
   6. Meeting intelligence
   7. Connected applications
   8. Credits and usage concepts
2. Responsible Business Use of AI
   1. AI capabilities and limitations
   2. Generative AI output risks
   3. Hallucinations and unsupported conclusions
   4. Human accountability
   5. Confidential and personal information
   6. Approved data sources
   7. Data minimization
   8. Organizational instructions
   9. Permissions and access boundaries
   10. Audit and review considerations
3. Working with the Cassidy Knowledge Base
   1. Knowledge Base structure
   2. Collections and folders
   3. Document preparation
   4. File and folder imports
   5. OneDrive and connected sources
   6. Source synchronization
   7. Knowledge selection
   8. Permission-controlled access
   9. Document verification
   10. Outdated and conflicting information
   11. Source-grounded responses
4. Using Cassidy Chat
   1. Agent selection
   2. File uploads
   3. Knowledge Base access
   4. Workflow execution
   5. Effective business requests
   6. Context and supporting information
   7. Response formatting
   8. Follow-up instructions
   9. Source checking
   10. Output refinement
5. Creating Effective AI Instructions
   1. Task definition
   2. Role and responsibility
   3. Required inputs
   4. Business terminology
   5. Decision criteria
   6. Rules and restrictions
   7. Output structure
   8. Tone and writing style
   9. Missing-information handling
   10. Escalation conditions
   11. Quality-control instructions
6. Building Cassidy Agents
   1. Agent creation with natural language
   2. Agent purpose and scope
   3. Knowledge Base grounding
   4. Global and agent-level instructions
   5. Tools and connectors
   6. Read and write permissions
   7. Conversation testing
   8. Boundary testing
   9. Agent sharing
   10. Agent maintenance
7. Insurance and Underwriting Applications
   1. Submission information extraction
   2. Document classification
   3. Missing-information identification
   4. Policy and guideline retrieval
   5. Risk-information summaries
   6. Underwriting file preparation
   7. Broker communication drafting
   8. Referral-note preparation
   9. Renewal comparison support
   10. Claims-history summarization
   11. Quality-assurance checklists
   12. Exception identification
   13. Human decision ownership
8. Creating Cassidy Workflows
   1. Workflow purpose and structure
   2. Workflow Copilot
   3. Templates and blank workflows
   4. Triggers
   5. Inputs
   6. AI actions
   7. Conditional paths
   8. Connected application actions
   9. Structured outputs
   10. Human-in-the-loop reviews
   11. Approval stages
   12. Notifications
   13. Publishing and execution
9. Underwriting Workflow Design
   1. Submission intake
   2. Document completeness checks
   3. Data extraction
   4. Information normalization
   5. Guideline lookup
   6. Risk-factor identification
   7. Missing-data requests
   8. Referral routing
   9. Summary generation
   10. Reviewer approval
   11. Record updates
   12. Workflow exception handling
10. Testing and Improving Results
    1. Test-case selection
    2. Expected-result definition
    3. Source verification
    4. Accuracy checking
    5. Consistency checking
    6. Instruction refinement
    7. Edge-case testing
    8. Permission testing
    9. Failure handling
    10. User feedback
    11. Workflow version control
    12. Ongoing performance review
11. Operational Adoption
    1. Suitable process selection
    2. Small-scale pilot design
    3. Process ownership
    4. User responsibilities
    5. Approval requirements
    6. Documentation standards
    7. Change management
    8. Usage monitoring
    9. Credit management
    10. Continuous improvement
    11. Governance and compliance alignment

Disclaimer

This course outline is provided as a general training guideline and does not constitute a fixed or binding programme of instruction. The trainer may amend, reorganize, replace, expand, or omit any topic, activity, or sequence when reasonably necessary to accommodate participant needs, organizational requirements, platform changes, time constraints, or professional judgment, without prior notice.