Getting Started with Kaman

Your Journey to Intelligent Automation Begins Here

Welcome to Kaman! This guide will walk you through the key concepts, help you understand what's possible, and show you how to begin transforming your operations with intelligent automation through our chat-first experience.

---

Understanding Kaman

What Kaman Does for Your Organization

Core Capabilities at a Glance

---

Chat-First Experience

Kaman's interface puts conversation at the center. Start simple and progressively discover more capabilities.

Progressive Disclosure

1. Chat First - Ask questions naturally
2. Suggestions - AI recommends next steps
3. Templates - Pre-built solutions for common needs
4. Sub-Agents - Specialized agents for complex tasks
5. Full Configuration - Advanced options when needed

---

Key Concepts

The AI Assistant

Your intelligent interface to Kaman. Ask questions in natural language, request actions, and get work done without navigating complex systems.

What You Can Ask:

• Questions about your data ("What were our top support issues last month?")
• Action requests ("Create a report on Q3 performance with charts")
• Process execution ("Start the customer onboarding workflow for Acme Corp")
• Complex analysis ("Analyze this data and identify anomalies")

What You'll See:

• Thinking indicators showing the AI's reasoning process
• Tool progress as actions are executed
• Sub-agent spawning for complex tasks
• Rich artifacts like charts, tables, and applications

Key Principle: The assistant works within your permission boundaries - it can only access what you're authorized to see.

Sub-Agents

For complex tasks, the main agent can spawn specialized sub-agents:

Workflows

Automated processes that execute business logic across multiple systems and steps.

Workflow Features:

• Triggers - What starts the workflow (event, schedule, or manual)
• Steps - Actions to perform
• Decisions - Conditional branching based on data
• Human-in-the-Loop - Pause for approvals when needed
• Checkpoints - Resume from any point after interruptions

5-Layer Intelligent Memory

More than storage - Kaman automatically analyzes your data to discover:

• Layer 0 (Redis) - Immediate context and working memory
• Layer 1-2 (Neo4j) - Entities, relationships, and knowledge graphs
• Layer 3-4 (PostgreSQL) - Historical conversations and semantic search

Discovers:

• Entities - People, companies, projects, products
• Relationships - How entities connect to each other
• Patterns - Recurring behaviors and trends
• Anomalies - Unusual activities worth investigating

Connectors & Channels

Connectors (23+ MCPs) link Kaman to your data sources:

• Databases (PostgreSQL, MySQL, MongoDB, Oracle, etc.)
• Cloud services (Salesforce, ServiceNow, Google, Zoho)
• File systems and code repositories (SFTP, GitHub, GitLab)

Channels (14) enable communication - see Channels documentation:

• Messaging: Slack, Teams, Mattermost, Discord, Telegram
• Business: WhatsApp (Business & Personal), Twilio SMS, Email, Gmail
• Automation: HTTP Webhook, Cron, Google Workspace, Odoo Events

---

Your First Steps

Step 1: Start with Chat

The main dashboard is chat-first. Simply type your question or request:

Step 2: Ask Your First Question

Open the AI Assistant and try a simple question:

Good starting questions:

• "What can you help me with?"
• "Show me what data sources are connected"
• "What workflows are available?"
• "Help me understand my recent activity"

Tips for effective questions:

• Be specific about what you need
• Include relevant context
• Specify time ranges when relevant
• Ask follow-up questions to refine results

Step 3: Explore Templates

Templates help you get started quickly:

• Application Templates - Pre-built AI assistants for specific use cases
• Workflow Templates - Automated processes ready to deploy
• Agent Templates - Specialized agents for domain tasks
• Data Lake Templates - Pre-configured analytics setups

Step 4: Try a Simple Task

Ask the AI to do something useful:

"Generate a summary of our support tickets from last week, identify the top 3 issues, and create a chart showing the trend"

Watch as the AI:

1. Shows its thinking process
2. Queries your connected data
3. Analyzes the results
4. Generates a chart artifact
5. Provides recommendations

Step 5: Connect Your Data

To unlock full capabilities, connect your data sources:

1. Navigate to Integrations
2. Choose from 23+ MCP connectors
3. Follow the guided setup
4. Configure sync settings
5. Start querying your data via chat

---

Best Practices

Working with the AI Assistant

Do:

• Be specific in your requests
• Provide context when helpful
• Review artifacts and verify important outputs
• Give feedback to improve suggestions
• Watch the thinking indicators to understand reasoning

Don't:

• Assume it knows implicit context
• Skip verification for critical decisions
• Share sensitive information unnecessarily
• Ignore the sub-agent notifications

Building Workflows

Start Simple:

• Begin with straightforward automations
• Test thoroughly before production use
• Use checkpoint recovery for long-running processes
• Plan for human-in-the-loop approvals

Scale Gradually:

• Add complexity incrementally
• Monitor with Observatory telemetry
• Gather user feedback
• Optimize based on results

Managing Data

Data Quality:

• Understand your data sources
• Monitor data quality in KDL
• Address issues promptly
• Keep documentation current

Access Control:

• Follow principle of least privilege
• Use multi-scope permissions appropriately
• Review access regularly
• Document access decisions

---

Getting Help

In-Platform Resources

• Help Center - Searchable documentation
• AI Assistant - Ask "How do I...?" questions
• Tooltips - Hover for contextual help
• Suggestions - AI-recommended next steps

Common Questions

Q: How do I connect a new data source? A: Go to Integrations > Add Connection. Choose from 23+ MCP connectors and follow the guided setup.

Q: Can I undo a workflow action? A: Workflows can be configured with approval steps and checkpoints. Use time-travel in KDL to see historical data states.

Q: How do I share insights with my team? A: Use the export feature, share artifacts directly, or set up scheduled reports with notification workflows.

Q: What if I need a feature that doesn't exist? A: Check the Template Marketplace for extensions, or contact support to discuss custom solutions.

Q: How do sub-agents work? A: For complex tasks, the main agent automatically spawns specialized sub-agents. You'll see notifications when this happens.

---

Next Steps

For Business Users

1. Explore the Chat Interface - Spend time asking questions naturally
2. Browse Templates - Find pre-built solutions for your use cases
3. Connect Key Data - Enable access to your important data sources
4. Learn Artifacts - Understand how to use generated charts and applications

For Administrators

1. Configure Security - Set up multi-scope permissions
2. Set Up Integrations - Connect data sources via MCPs
3. Define Policies - Establish access controls and governance
4. Monitor Operations - Use Observatory for visibility

For Power Users

1. Master Workflows - Learn advanced workflow automation
2. Leverage Memory - Explore 5-layer memory capabilities
3. Build with Templates - Create custom templates for your team
4. Optimize Processes - Find automation opportunities

---

Quick Reference

Keyboard Shortcuts

Key Terms

---

Ready to Begin?

You now have the foundation to start using Kaman effectively. Remember:

• Start with chat - The AI understands natural language
• Watch the thinking - See how the AI reasons through requests
• Explore gradually - Discover more features as you need them
• Ask for help - The AI Assistant and support team are here for you
• Share feedback - Your input helps improve the platform

Welcome to intelligent automation!

---

Getting Started - Your foundation for success with Kaman