Overview - NeuroAI

The Complete Agent System

NeuroAI uses a powerful five-layer approach to create intelligent, specialized agents:

Layer 1: Custom Agents (WHAT to do)

Define the role, responsibilities, and core functions

Layer 2: Personas (HOW to work)

Set expertise, communication style, and work standards

Layer 3: Tools (WHICH capabilities to use)

Enable specific technical capabilities (29+ tools available)

Layer 4: Apps (WHERE to access data)

Connect to 500+ external services and data sources

Layer 5: Knowledge Base (WHAT to know)

Provide institutional knowledge and company context Result: Specialized AI workers that understand your business, follow your processes, use the right tools, access your data, and deliver consistent results.

Layer 1: Custom Agents

What is a Custom Agent?

A Custom Agent is a specialized AI worker you create for specific roles or tasks. It’s the world’s first one-prompt agent creation platform—describe what you need, and your agent is ready instantly.

Key Features

One-Prompt Creation

Single Request:

"Create an AI agent that generates weekly industry reports 
on AI funding, product launches, and regulatory changes"

Result:

Agent created instantly
Ready to use immediately
Reusable across conversations
Shareable with team

Reusability

Create Once, Use Forever:

Create: “Create a competitive intelligence agent”
Invoke: Type @ in conversation to call it
Reuse: Same agent works across all chats
Share: Publish to team or Agent Store

Collaboration (Coming Soon)

Multi-Agent Workflows:

@MarketResearcher + @FinancialAnalyst + @ContentWriter

→ Researcher gathers market data
→ Analyst creates projections
→ Writer synthesizes into report

Result: Complex tasks through agent collaboration

Types of Custom Agents

1. Role-Based Agents

Agents that embody specific professional roles:

Business Strategist → Strategic planning and competitive analysis
Market Analyst → Market research and trend analysis
Financial Analyst → Financial modeling and projections
Product Manager → Product development and roadmaps
Content Creator → Marketing and social content

2. Industry-Specific Agents

Agents tailored to specific industries:

Healthcare AI Agent → HIPAA compliance, medical terminology, healthcare business models
SaaS Business Agent → SaaS metrics (CAC, LTV, MRR), subscription economics
Real Estate Agent → Property market analysis, investment evaluation
Startup Agent → Early-stage business expertise, venture funding landscape

3. Task-Specific Agents

Agents focused on specific tasks:

Competitive Intelligence Officer → Monitors competitors, analyzes strategic implications
Research Curator → Aggregates industry content, curates relevant articles
Report Generator → Creates formatted reports, data visualization
Social Media Manager → Content creation, engagement tracking, trend monitoring

Creating Custom Agents

Creation Process

Step 1: Access Creation

Homepage: Click “Custom Agent”
Sidebar: Click ”+ New” → “Custom Agent”
In-Chat: Type @ → “Create New Agent” (coming soon)

Step 2: Define Your Agent Use Structured Template:

[Role Definition] + [Core Functions] + [Output Standards]

Example:
"Create an AI industry analyst that generates weekly 
reports on funding trends, technology developments, 
and policy interpretations for investment decision-making"

Step 3: Test in Preview

Test agent in Preview window
Review output quality
Adjust instructions as needed
Click “Update” to save

Step 4: Start Using

Agent appears in your library
Invoke with @ mention
Use across conversations
Share with team if needed

Using Custom Agents

Invocation Methods

@Mention in Conversation:

In Super Agent chat:
"@CompetitiveIntelligence analyze our top 3 competitors"

Result:
→ Specific agent handles the task
→ Uses its specialized expertise
→ Delivers in its configured style

Multi-Agent Collaboration (Coming Soon):

Complex Task: "Create an investor pitch deck"

Workflow:
1. @MarketResearcher analyzes market
2. @CompetitiveAnalyst researches competitors
3. @FinancialAnalyst creates projections
4. @PitchDeckDesigner builds presentation

Result: Collaborative AI team delivers complete deck

Chained Workflows:

"@ResearchAgent find latest AI trends, then 
@ContentWriter create a blog post about them, then
@SocialMediaManager create posts promoting it"

Result: End-to-end content workflow

Managing Custom Agents

Edit Agent:

Click three-dot menu on agent card
Select “Configure”
Modify through conversation or manual edits
Click “Update” to save changes

Delete Agent:

Click three-dot menu
Select “Delete”
Confirm deletion (permanent and irreversible)

Share Agent: Sharing Options:

Only me (private)
Anyone with the link (team sharing)
Custom Agent Store (public)

How to Share:

Click three-dot menu
Select “Configure”
Click “Share” button
Choose sharing scope

Custom Agent Store

Discover Agents: Browse Categories:

Writing
Productivity
Research & Analysis
Education
Lifestyle
Others

Filter Options:

My Super Agents → Agents you created or favorited
Trending → Sorted by favorites count
By Genspark → Officially published agents

Using Store Agents:

Browse Custom Agent Store
Click agent card to view details
Favorite to add to “My Super Agents”
Invoke via @ mention in conversations

Layer 2: Personas

What is a Persona?

A Persona is a customizable instruction set that defines how an agent behaves, communicates, and operates. It’s like a detailed job description that shapes the agent’s expertise, style, and work standards.

Why Use Personas?

The Problem with Generic AI

Generic AI Assistant:

✗ Same style for everyone
✗ Same detail level for all tasks
✗ Generic assumptions
✗ One-size-fits-all output
✗ No specialization

The Persona Solution

Persona-Based Agent:

✓ Consistent behavior aligned with your needs
✓ Domain expertise in specific areas
✓ Communication style matching your brand
✓ Work preferences following your processes
✓ Quality control to your standards

Persona Components

1. Core Purpose

Example:
"Create executive-level strategic plans and business analyses"

2. Expertise Areas

Expert Level:

Business strategy
Market analysis
Financial planning
Competitive intelligence

Proficient Level:

International markets
Emerging technologies
Industry trends

Basic Level:

Regulatory requirements
Local variations

3. Communication Style

Style Definition:

Tone: Professional but approachable
Detail: Executive-friendly
Format: Concise and direct
Approach: Data-driven and action-oriented

4. Output Format

Structure:

Executive summary first
Key metrics highlighted
Recommendations clearly stated
No jargon or unnecessary detail

5. Work Preferences

Process:

Start with executive summary
Provide citations for all data
Use at least 3 sources per claim
Highlight key metrics in tables
End with actionable recommendations
Flag assumptions and limitations

6. Quality Standards

Requirements:

Zero factual errors
All statistics cited
Sources verified
Assumptions clearly stated
Limitations acknowledged

Creating Personas

Step 1: Define Purpose

What will this Persona do? ✅ Specific:

"Create detailed market research presentations 
for enterprise clients"

❌ Vague:

"Help with presentations"

Step 2: Define Expertise

What should this Persona know deeply? Example: Market Research Persona Expertise Areas:

Market sizing and analysis
Competitive intelligence
Industry trends and forecasting
Customer research methodologies
Data visualization
Financial analysis of opportunities

Step 3: Set Communication Style

How should this Persona communicate?

Tone: Professional, data-driven, executive-friendly
Formality: Formal (not casual)
Detail Level: Comprehensive but organized
Jargon: Use industry terms (technical audience)
Length: Concise summaries with supporting detail
Format: Clear sections with key takeaways

Step 4: Define Work Preferences

How should this Persona operate? Process:

Always start with executive summary
Provide citations for all data points
Include at least 3 data sources per claim
Highlight key metrics in tables
End with actionable recommendations
Flag assumptions and limitations
Use consistent formatting
Prioritize accuracy over speed

Step 5: Add Constraints

What are the rules this Persona must follow? Constraints:

Research only 2024-2025 data
Avoid speculation without caveats
Always cite sources with links
Flag potential conflicts of interest
Avoid jargon without explanation
Keep presentations to max 20 slides
Ensure business-appropriate visuals

Step 6: Define Success Metrics

How do you measure success? Success Criteria:

✓ Minimal editing required
✓ Consistent formatting
✓ Accurate, cited information
✓ 30% faster than standard agent
✓ Executives approve without revision
✓ Maintains quality across all outputs

Persona Examples

Example 1: Executive Strategist

Name: “Executive Strategist” Purpose:

"Create strategic analyses and business plans 
for global enterprise market"

Expertise:

Enterprise software dynamics
B2B sales strategies
Global market expansion
Risk management
Fortune 500 expectations

Communication:

Executive-first (C-suite audience)
Data-intensive (every claim cited)
Conservative (risks highlighted)
Professional formal tone
Global perspective

Output Standards:

10-15 page analyses
3-5 page executive summary first
Detailed competitive analysis
Risk assessment included
Clear recommendations with timelines
Investment case included

Success Metrics:

Board-ready quality
Minimal revisions needed
Decision-enabling output

Example 2: Startup Fundraiser

Name: “Founder’s Strategist” Purpose:

"Support fundraising, product-market fit, 
and rapid scaling"

Expertise:

Venture capital landscape
Startup metrics and benchmarks
Product-market fit validation
Go-to-market for startups
Pitch deck creation

Communication:

Confident and ambitious
Data-driven but opportunity-focused
Clear and concise
Action-oriented
Forward-looking

Output Standards:

Pitch decks (12-15 slides, VC-ready)
Market sizing analyses
Competitive positioning
Fundraising strategy guides
Investor messaging templates

Success Metrics:

Gets meetings with target VCs
Investor updates generate interest
Market analyses defend valuations

Example 3: Healthcare Strategist

Name: “Healthcare Innovation Strategist” Purpose:

"Navigate healthcare market complexities 
and regulations"

Expertise:

Healthcare regulations (FDA, CMS, HIPAA)
Provider and payer incentives
Reimbursement models
Clinical validation
Digital health adoption

Communication:

Clinical accuracy required
Business-savvy
Regulatory awareness
Implementation-focused
Risk-aware

Domain Knowledge:

ICD-10 and CPT coding
EHR integration
Clinical workflows
Healthcare IT infrastructure
Medical device classification

Success Metrics:

Healthcare executives find credible
FDA/CMS pathways clearly articulated
Clinical team approves claims
Realistic reimbursement strategy

Combining Custom Agents + Personas

The Power of Integration

When you combine Custom Agents with Personas, you get specialized AI workers that:

Custom Agent: WHAT to do
  "Competitive Intelligence Officer"
  → Monitors competitors
  → Analyzes market changes
  → Delivers intelligence reports

+

Persona: HOW to do it
  "Strategic Analyst"
  → Data-driven approach
  → Executive-friendly format
  → Strategic implications focus
  → Weekly reporting cadence

=

Specialized AI Worker
  A Competitive Intelligence Officer that delivers
  strategic, data-driven competitor analysis in
  executive-friendly format every week

Real-World Integration Example

Sales Enablement Agent CUSTOM AGENT: “Sales Enablement Manager” Role:

Create sales materials
Develop battle cards
Write objection handlers
Design pitch decks

+ PERSONA: “Sales Strategist” Expertise:

Sales messaging
Competitive positioning
Customer pain points
ROI calculation

Communication:

Benefit-focused
Clear and persuasive
Action-oriented
ROI-centered

Output Format:

Sales decks
One-pagers
Battle cards
Email templates

= SPECIALIZED WORKER A Sales Enablement Manager that creates benefit-focused, ROI-driven sales materials with competitive positioning and objection handling, delivered in ready-to-use formats

Layer 3: Tool Configuration

What is Tool Configuration?

Tool Configuration allows you to enable, disable, and manage individual capabilities for your agents. Choose exactly which tools each agent needs, nothing more, nothing less. This ensures:

Focused agents — Each agent has only needed capabilities
Security — No unnecessary access to external services
Performance — Faster execution with fewer tools
Cost control — Pay only for what you use
29 tools available — Enable what you need (soon will be more)

Available Tools (16 Categories)

1. Task Management

Organize work with structured to-do lists All 5 capabilities enabled:

Clear All → Remove all tasks and sections completely
Create Tasks → Create organized task lists with sections
Delete Tasks → Remove tasks or sections
Update Tasks → Mark tasks complete or modify content
View Tasks → See all tasks, sections, and progress

2. Files & Folders

Create, edit, read, and organize files All 5 capabilities enabled:

Create File → Create new file with content
Delete File → Remove files (requires user confirmation)
Edit File → Make targeted edits to existing files
Full File Rewrite → Completely replace file content
Str Replace → Replace specific text in file

3. Web Search & Research

Find and analyze information online 2 / 2 capabilities enabled:

Web Search → Search internet for information (batch searches supported)
Scrape Webpage → Extract full text from multiple webpages

4. File Reader & Knowledge Base

Read and search documents 4 capabilities enabled:

Read File → Extract text from workspace files (supports batch mode)
Search File → Semantic search within documents
Global KB Sync → Download knowledge base files locally
Semantic Search → Natural language search across files

5. Web Browser

Browse websites, click buttons, fill forms All 4 capabilities enabled:

Browser Act → Perform browser actions via natural language
Browser Extract Content → Extract structured content from page
Browser Navigate To → Go to specific URL
Browser Screenshot → Take screenshot of current page

6. Presentations

Create and manage stunning slides All 8 capabilities enabled:

Create Slide → Create or update single slide (parallel execution)
Load Template Design → Reference design from presentation template
Export Presentation → Export to PPTX and PDF simultaneously
Delete Presentation → Remove entire presentation
Delete Slide → Remove specific slide
List Presentations → View all presentations
List Slides → View slides in presentation

7. Image Processing

Generate, edit, and analyze images 3 capabilities enabled:

Image Vision → View and analyze images
Generate Media → Create, edit, upscale, or remove backgrounds
Canvas Editor → Create interactive canvases for design

8. Automation & Triggers

Set up automatic execution All 7 capabilities enabled:

Create Scheduled Trigger → Schedule agent to run automatically
Create Event Trigger → Run agent on specific events
Toggle Scheduled Trigger → Enable/disable scheduled tasks
Delete Scheduled Trigger → Remove scheduled automation
Get Scheduled Tasks → View all scheduled runs

9. Credentials & Authentication

Manage API keys and integrations All 4 capabilities enabled:

Create Credential Profile → Create authentication for toolkit
Configure Profile For Agent → Add authenticated profile to agent
Get Credential Profiles → View existing profiles
Delete Credential Profile → Remove credential profile

10. Research Tools

Find information about people, companies, academic research 3 capabilities enabled:

People Research → Find professional background information
Company Research → Find business information
Academic Research → Search academic papers and researchers (FREE using Semantic Scholar)

11. Voice Calls

Make and manage phone calls All 5 capabilities enabled:

Make Phone Call → Initiate outbound call with AI agent
Wait For Call Completion → Monitor call until complete
Get Call Details → View call transcripts
End Call → Terminate active call
List Calls → View call history

12. Knowledge Base

Store and retrieve information All 9 capabilities enabled:

Global KB Upload File → Add file to knowledge base
Global KB Create Folder → Create folder for organization
List Contents → View knowledge base items
Enable Item → Enable knowledge base item
Delete Item → Remove from knowledge base

13. Agent Configuration

Modify agent settings 2 / 2 capabilities enabled:

Get Current Agent Config → Check current agent configuration
Update Agent → Modify agent configuration

14. Chat & Messages

Communicate with users 1 capability enabled:

Ask → Request user input or confirmation (with follow-up options support)

Tool Configuration Best Practices

✅ Do’s

✓ Enable only needed tools (focused agents are more reliable)
✓ Check current config before updating (use get_current_agent_config first)
✓ Request user confirmation for destructive actions
✓ Use batch operations (faster than individual calls)
✓ Monitor long operations (use wait functions)
✓ Provide clickable options (follow_up_answers when asking users)
✓ Attach files when sharing results
✓ Mark tasks complete immediately after each task finishes

❌ Don’ts

✗ Don’t enable unnecessary tools (adds complexity, increases costs)
✗ Don’t delete without confirmation (always ask user first)
✗ Don’t use old parameter names (use EXACTLY the specified names)
✗ Don’t pass JSON strings for arrays (pass actual array objects)
✗ Don’t run canvas operations in parallel (one at a time only)
✗ Don’t skip extraction after Apify runs
✗ Don’t assume file indexing complete (use ls_kb to check status)
✗ Don’t proceed without user confirmation for destructive operations

Common Configuration Scenarios

Scenario 1: Research Agent

Enabled tools:

✓ Web Search
✓ Scrape Webpage
✓ File Reader
✓ Semantic Search
✓ Task Management
✗ Browser Act (not needed)
✗ Voice Calls (not needed)

Capabilities: Search, read, analyze research

Scenario 2: Content Creation Agent

Enabled tools:

✓ Presentations
✓ Image Generation
✓ Canvas Editor
✓ Web Search
✓ File Creation
✓ Task Management
✓ Web Browser (for research)

Capabilities: Create content, presentations, images

Scenario 3: Automation Agent

Enabled tools:

✓ Scheduled Triggers
✓ Event Triggers
✓ Web Browser
✓ Web Search
✓ Task Management
✓ Credentials Manager
✓ File Creation

Capabilities: Automate workflows, scheduled execution

Scenario 4: Business Research Agent

Enabled tools:

✓ People Research
✓ Company Research
✓ Web Search
✓ File Reader
✓ Presentations
✓ Task Management

Capabilities: Research businesses and people

Layer 4: Apps Integration

What are Apps?

Apps enable your custom agents to access external tools, search data, and take actions across your entire tech stack. Instead of being limited to conversation, agents can:

Search — Find information from connected services
Reference — Pull real-time data into responses
Take Actions — Create tasks, send messages, update records
Deep Research — Analyze multiple sources with citations
Integrate — Connect to 500+ business tools

How Apps Work With Custom Agents

Agent Capabilities With Apps

When you configure apps for a custom agent, it gains access to those services: Agent Request:

"Search our CRM for accounts in healthcare"

With Apps Enabled:

→ Agent accesses Salesforce via app
→ Searches for matching accounts
→ Returns results with relevant data

Without Apps:

→ Agent responds with generic knowledge
→ No access to actual company data
→ Limited to training data

Agent Personas & Their Apps

Different agent personas benefit from different app combinations:

1. Sales Agent

Purpose: Drive revenue through intelligent sales outreach and lead management Core Apps:

Salesforce — Access accounts, leads, opportunities
Apollo — Find and verify prospect emails
LinkedIn — Research prospects and companies
Gmail — Send personalized emails
Slack — Update team on progress
HubSpot — Manage contacts and deals
Pipedrive — Track sales pipeline

Key Workflows:

Lead Research
- App: Apollo + LinkedIn
- → Find prospects
- → Research background
- → Verify email addresses
Account Management
- App: Salesforce + HubSpot
- → Pull account information
- → Check recent activities
- → Track deal status
Email Outreach
- App: Gmail + Salesforce
- → Draft personalized email
- → Send via Gmail
- → Log to Salesforce
Team Updates
- App: Slack + Salesforce
- → Pull yesterday’s closed deals
- → Post summary to #sales
- → Notify team of updates

Agent Capabilities:

✓ Search Salesforce for specific accounts
✓ Find prospect emails with Apollo
✓ Draft personalized outreach emails
✓ Log activities to CRM
✓ Update deal status
✓ Post team updates to Slack
✓ Research companies on LinkedIn
✓ Track pipeline progress

2. Customer Success Agent

Purpose: Ensure customer satisfaction and drive retention through proactive support Core Apps:

Salesforce — View customer accounts and history
Zendesk — Manage support tickets
Intercom — Send customer messages
Google Drive — Access success plans
Slack — Notify team members
HubSpot — Track customer interactions
Freshdesk — Manage support cases

Key Workflows:

Onboarding Support
- App: Salesforce + Zendesk + Google Drive
- → Pull onboarding checklist from Drive
- → Create support tickets in Zendesk
- → Log in Salesforce
- → Send welcome message via Intercom
Health Monitoring
- App: Salesforce + Intercom
- → Check customer health score
- → Review recent activity
- → Send proactive outreach
- → Log interaction to account
Issue Resolution
- App: Zendesk + Slack
- → Pull ticket details
- → Research solution
- → Update ticket status
- → Notify team in Slack

3. Marketing Agent

Purpose: Generate demand and enhance brand awareness through intelligent campaigns Core Apps:

Mailchimp — Email marketing campaigns
HubSpot — Lead generation and tracking
Google Analytics — Website metrics
Canva — Design graphics
LinkedIn — Post content
Facebook/Instagram — Social posting
Google Drive — Access marketing assets
Slack — Share updates

Key Workflows:

Campaign Planning
- App: HubSpot + Google Drive
- → Review past campaign performance
- → Pull creative assets
- → Plan new campaign
- → Document strategy
Email Campaign
- App: Mailchimp + HubSpot
- → Create email template
- → Define audience segment
- → Schedule send
- → Track performance
Content Creation
- App: Canva + Google Drive + LinkedIn
- → Design social graphics
- → Create LinkedIn post
- → Schedule publication
- → Track engagement

4. Product Manager Agent

Purpose: Drive product development with data-driven decisions and customer insights Core Apps:

Jira — Track features and bugs
Linear — Issue management
Notion — Product roadmap
Intercom — Customer feedback
Google Drive — Design docs
GitHub — Code repository
Slack — Team communication
Mixpanel — Product analytics

Key Workflows:

Roadmap Planning
- App: Notion + Jira + Intercom
- → Pull customer feedback
- → Review feature requests
- → Create roadmap items
- → Prioritize based on impact
Feature Tracking
- App: Jira + Linear + Slack
- → Create feature spec
- → Break into tasks
- → Assign to engineering
- → Track progress

5. Operations Agent

Purpose: Streamline processes and ensure organizational efficiency Core Apps:

Asana — Project management
Monday.com — Task tracking
Google Sheets — Data management
Slack — Team communication
Google Drive — Document storage
ClickUp — Workflow management
Zapier — Automation
Google Calendar — Scheduling

Key Workflows:

Project Management
- App: Asana + Slack
- → Create project tasks
- → Assign to team
- → Track progress
- → Update team in Slack
Process Automation
- App: Zapier + Google Sheets
- → Create automation workflows
- → Track execution
- → Log data to sheets
- → Monitor performance

6. Research & Analytics Agent

Purpose: Provide data-driven insights and competitive intelligence Core Apps:

Perplexity — Deep research
Google Analytics — Website data
Firecrawl — Web scraping
Tavily — Information retrieval
Exa — Semantic search
Bright Data — Market data
Notion — Documentation
Google Sheets — Data compilation

Key Workflows:

Market Research
- App: Perplexity + Firecrawl
- → Research market trends
- → Gather competitor info
- → Compile statistics
- → Document findings
Competitive Intelligence
- App: Bright Data + Tavily + Notion
- → Track competitor activities
- → Monitor pricing changes
- → Research product updates
- → Create competitive analysis

7. Content Creator Agent

Purpose: Produce high-quality content efficiently across multiple channels Core Apps:

Canva — Design graphics
Google Drive — Store content
Figma — Create designs
LinkedIn — Post content
YouTube — Video management
HeyGen — Video creation
Notion — Content calendar
Slack — Share updates

Key Workflows:

Social Content
- App: Canva + LinkedIn
- → Design social graphic
- → Write caption
- → Schedule post
- → Track engagement
Blog Content
- App: Google Drive + Notion
- → Create outline
- → Write article
- → Add visuals
- → Schedule publication

8. HR & Recruitment Agent

Purpose: Streamline hiring and employee management processes Core Apps:

Lever — Applicant tracking
BambooHR — Employee data
Google Drive — Documents
Slack — Team communication
Gmail — Email communication
Google Calendar — Interview scheduling
Workable — Recruitment
Calendly — Scheduling

Key Workflows:

Recruitment
- App: Lever + Gmail
- → Post job opening
- → Screen applications
- → Send interview invites
- → Track candidates
Onboarding
- App: BambooHR + Google Drive
- → Create employee record
- → Prepare onboarding docs
- → Schedule training
- → Track completion

9. Finance Agent

Purpose: Manage financial processes and provide business insights Core Apps:

Stripe — Payment processing
Xero — Accounting
QuickBooks — Financial management
Google Sheets — Data tracking
Slack — Team updates
Google Drive — Document storage
Tableau — Reporting
Square — POS/Payments

Key Workflows:

Invoice Management
- App: Xero + Stripe + Google Sheets
- → Create invoices
- → Track payments
- → Send reminders
- → Log in spreadsheet
Financial Reporting
- App: QuickBooks + Google Sheets + Slack
- → Pull financial data
- → Create reports
- → Analyze trends
- → Post summaries

10. IT & DevOps Agent

Purpose: Manage infrastructure and ensure system reliability Core Apps:

GitHub — Code repository
Sentry — Error tracking
DataDog — Monitoring
Slack — Alerts and communication
Jira — Issue tracking
CloudFlare — Security
New Relic — APM
Vercel — Deployment

Key Workflows:

Incident Management
- App: Sentry + DataDog + Slack
- → Monitor errors
- → Create alerts
- → Notify team
- → Track resolution
Deployment Management
- App: GitHub + Vercel + Slack
- → Track deployments
- → Monitor performance
- → Rollback if needed
- → Post updates

Configuring Apps for Your Agent

Step 1: Identify Agent Needs

Ask yourself:

What tools does this agent need?
What data should it access?
What actions should it take?
What notifications are needed?

Step 2: Select Relevant Apps

For each app:

Verify the agent needs it
Check what permissions are required
Ensure data is appropriate
Plan the workflows

Step 3: Configure Permissions

For each app:

Grant read permissions if searching
Grant write permissions if creating/updating
Restrict access to specific folders/teams
Document what access is granted

Step 4: Test Agent Workflows

Before deploying:

Test basic app access
Verify search capabilities
Test write actions
Confirm notifications work
Review output quality

Step 5: Deploy and Monitor

After deployment:

Monitor app usage
Check for errors
Gather feedback
Refine permissions
Scale as needed

App Permissions & Security

Permission Levels

READ Permissions:

✓ Search documents
✓ View records
✓ Access data
✗ Create or modify
✗ Delete
✗ Share with others

WRITE Permissions:

✓ All READ actions
✓ Create new records
✓ Update existing data
✓ Delete if enabled
✗ Share or change permissions
✗ Manage team access

ADMIN Permissions:

✓ All READ and WRITE actions
✓ Manage team access
✓ Change permissions
✓ Configure settings

Best Practices

✅ Grant minimal necessary permissions
✅ Review permissions quarterly
✅ Use specific folder/team restrictions
✅ Document what each app accesses
✅ Revoke unused app permissions
✅ Audit app activity regularly
✅ Test with sandbox first
✅ Monitor for unusual access

Layer 5: Knowledge Base

What is a Knowledge Base?

A Knowledge Base is a custom repository of information that informs how your Neuro agents work. Instead of using generic knowledge, your agents reference your Knowledge Base to understand your business, preferences, processes, and expertise. Think of a Knowledge Base as the “institutional library” for your AI team. It contains everything your agents need to know to work effectively: company values, industry expertise, best practices, standard operating procedures, brand guidelines, and customer information.

Why Build a Knowledge Base?

The Problem Without a Knowledge Base

Without a Knowledge Base, your agents:

✗ Don’t know your company’s specific processes
✗ Don’t understand your industry expertise
✗ Can’t maintain consistent brand voice
✗ Make generic recommendations
✗ Require detailed explanations for every task
✗ Miss context about your business

The Knowledge Base Solution

With a Knowledge Base, your agents:

✓ Understand your business — Know your market, products, and strategy
✓ Follow your processes — Execute work your way consistently
✓ Maintain your brand — Write and communicate like your company
✓ Provide relevant insights — Make recommendations based on your context
✓ Work independently — Need less explanation per task
✓ Improve over time — Learn from your feedback and updates

What Goes in a Knowledge Base?

Industry Knowledge

Market Trends & Insights

"AI automation market trends (2025):
- 340% growth in enterprise automation adoption
- 60% of enterprises using 3+ automation tools
- Shift from RPA to AI-native platforms
- Increasing focus on cost reduction

Market opportunity:
- $65B+ global market
- Growing 32% CAGR
- Consolidation expected
- Significant vertical integration potential"

Regulatory & Compliance

"Key regulations affecting our business:
- GDPR (EU data protection)
- CCPA (California privacy)
- SOC 2 Type II (enterprise requirement)
- Industry-specific: HIPAA, PCI-DSS for specific verticals

Our compliance posture:
- GDPR compliant
- SOC 2 Type II certified
- CCPA compliant
- Custom NDA support"

Brand & Communication Guidelines

Brand Voice & Tone

"Our brand voice:
- Professional but approachable
- Data-driven but human-centered
- Confident without arrogance
- Educational, not condescending
- Action-oriented and pragmatic

Communication style:
- Concise and clear
- Avoid jargon where possible
- Use examples over abstractions
- Focus on customer value, not features"

Messaging Framework

"Core messaging:
1. Problem: Enterprise software automation is hard
2. Solution: Our AI platform makes it simple
3. Value: Save 10+ hours per employee per week
4. Proof: 500+ customers, 40% growth rate

Sales messaging by segment:
- SMB: Cost savings and time savings
- Enterprise: Risk reduction and compliance
- Developers: Flexibility and customization"

Process Documentation

Standard Operating Procedures

"Sales process:
1. Discovery call (30 min)
   - Understand customer needs
   - Assess fit (80%+ fit required)
   
2. Product demo (45 min)
   - Show relevant features
   - Answer questions
   
3. Trial setup (1 week)
   - Free trial environment
   - Onboarding support
   
4. Close (2-4 weeks)
   - ROI presentation
   - Contract negotiation
   - Implementation planning"

Customer Onboarding

"Standard onboarding flow:
- Week 1: Project kickoff, requirements gathering
- Week 2: Platform training and customization
- Week 3: Testing and quality assurance
- Week 4: Launch and support hand-off

Metrics for success:
- 90% of use cases implemented
- Customer achieves ROI within 90 days
- Adoption at 70%+ of team
- Customer satisfaction score 8/10+"

Building Your Knowledge Base: Step-by-Step

Step 1: Assess What You Know

Before building your Knowledge Base, inventory what information exists: Information Audit: Strategic Information:

☐ Company mission, vision, values
☐ Business strategy and goals
☐ Competitive analysis
☐ Market positioning
☐ 3-year plan

Product Information:

☐ Product strategy and roadmap
☐ Feature documentation
☐ Customer use cases
☐ Pricing model
☐ Integration documentation

Operational Information:

☐ Sales process and playbooks
☐ Onboarding procedures
☐ Customer success metrics
☐ Support processes
☐ Standard templates

Brand Information:

☐ Brand guidelines
☐ Logo and visual standards
☐ Messaging framework
☐ Company voice and tone
☐ Communication standards

Step 2: Organize by Category

Structure your Knowledge Base logically:

Knowledge Base Structure:

├── Company
│   ├── Mission & Values
│   ├── Strategy
│   ├── Business Model
│   ├── Competitive Landscape
│   └── Organization
│
├── Products & Services
│   ├── Product Overview
│   ├── Roadmap
│   ├── Features & Capabilities
│   ├── Pricing Model
│   └── Integrations
│
├── Industry & Market
│   ├── Market Trends
│   ├── Customer Research
│   ├── Competitive Intelligence
│   ├── Regulatory Requirements
│   └── Best Practices
│
├── Brand & Communication
│   ├── Brand Guidelines
│   ├── Messaging Framework
│   ├── Voice & Tone
│   ├── Visual Standards
│   └── Templates
│
├── Processes & Operations
│   ├── Sales Process
│   ├── Customer Onboarding
│   ├── Support Procedures
│   ├── Content Standards
│   └── Decision Framework
│
└── Reference Materials
    ├── Case Studies
    ├── Customer Data
    ├── Technical Documentation
    ├── Research & Reports
    └── Training Materials

Step 3: Document Core Information

Start with the most important, foundational information: Priority 1: Critical Business Context

Company mission and values
Target customer profile
Core value proposition
Competitive differentiation
Key business metrics

Priority 2: Essential Processes

Sales process and cycle
Customer onboarding
Support procedures
Decision-making process
Budget allocation

Priority 3: Brand & Communication

Brand guidelines
Tone of voice
Messaging framework
Key talking points
Communication standards

Priority 4: Reference Materials

Competitive intelligence
Market data
Customer insights
Technical documentation
Case studies

Step 4: Add to Neuro

Upload documents to your Neuro Knowledge Base: Upload Options:

Direct Text Input
- Paste document content directly
- Good for short, focused documents
Document Upload
- Upload PDF, Word, or text files
- Good for existing documentation
Link Integration
- Connect Google Drive, Notion, Confluence
- Automatic synchronization
- Keep content fresh
Manual Entry
- Create documents in Neuro interface
- Good for new content

Knowledge Base Best Practices

1. Keep Information Current ✅

Update Schedule:

Strategic Information: Quarterly
Operational Information: Monthly
Product Information: As-needed
Market & Research: Continuously

2. Make Information Specific ✅

❌ Too Vague:

"We believe in customer success"
"Our product is best-in-class"
"We compete on quality"

✅ Specific:

"Customer success means 90% of customers achieve 
their stated ROI goal within 90 days. We measure 
this through quarterly business reviews and track 
against our NPS of 65+."

"Our product is best-in-class because:
- 40% faster implementation than Zapier
- 10x cheaper than UiPath
- 3x more powerful than Make
- Built for enterprise reliability"

3. Include Examples ✅

Without Examples:

"Our sales process focuses on understanding 
customer needs and demonstrating product fit."

With Examples:

"Our sales process:
1. Discovery call (30 min)
   Example: 'Tell me about your biggest operational bottleneck'
   
2. Product demo (45 min)
   Example: Show how customers automated invoice processing
   
3. Trial (1 week)
   Example: Customers typically process 1000+ documents
   
4. Close (2-4 weeks)
   Example: ROI presentation shows $200K+ annual savings"

4. Cross-Reference Documents ✅

Linking Strategy: Document: “Sales Process” Links to:

→ Messaging Framework (what to say)
→ Company Positioning (why we’re different)
→ Competitive Intelligence (competitive talking points)
→ Customer Success Metrics (what success looks like)
→ Contract Templates (standard terms)

5. Use Consistent Terminology ✅

Define & Use Consistently:

Term: "Enterprise Customer"
Definition: "Companies with 1000+ employees 
and $500M+ revenue"

Term: "Mid-Market Customer"
Definition: "Companies with 100-1000 employees 
and $20M-$500M revenue"

Term: "SMB Customer"
Definition: "Companies with <100 employees 
and <$20M revenue"

Use these exact definitions consistently throughout your Knowledge Base and agent communications.

Integration with Personas

How Knowledge Base Supports Personas

Persona: “Enterprise Sales Strategist” Knowledge Base Documents Used:

Company positioning (what to say)
Enterprise customer profile (who to talk to)
Sales process (how to sell)
Competitive intelligence (how we differentiate)
Messaging framework (specific language)
Case studies (proof points)
ROI framework (how to calculate value)

Result: Agent can create enterprise sales strategies without manual briefing, using company-specific context from Knowledge Base.

Layer 6: Scheduled Tasks (Automation)

What are Scheduled Tasks?

Scheduled Tasks let you automate recurring work by having Neuro agents execute tasks on a predefined schedule. Set it once, configure it properly, and your agents handle it automatically—daily reports, weekly research, monthly analysis, or any recurring workflow. Think of Scheduled Tasks as automated assistants that run on your calendar. Instead of manually running the same task weekly or daily, your agents execute it automatically at the exact time you specify.

Why Use Scheduled Tasks?

The Problem with Manual Recurring Work

Without Scheduled Tasks, recurring work means:

✗ Remember to run the same task daily or weekly
✗ Manually trigger the same workflow repeatedly
✗ Risk forgetting to run important tasks
✗ Inconsistent timing and quality
✗ Distraction from important work

The Scheduled Tasks Solution

With Scheduled Tasks, you get:

✓ Automation — Tasks run automatically at set times
✓ Consistency — Same time, same quality, every execution
✓ Reliability — Never miss a deadline or update
✓ Time savings — Reclaim hours per week
✓ Focus — Stop thinking about repetitive work

Quick Start: Three Ways to Use Scheduled Tasks

One-Time Task (Tomorrow)

Ask Neuro:

"Run this task tomorrow at 9 AM: 
Research top AI news from the past 24 hours 
and email me a summary"

When to use: Tasks that need to happen later today or tomorrow, but not recurring.

Recurring Task (Weekly)

Ask Neuro:

"Every Monday at 8 AM, research our top 5 competitors 
and send me a summary of any product updates or news 
from the past week"

When to use: Regular weekly tasks like competitor monitoring, report generation, or updates.

Complex Schedule (Monthly)

Ask Neuro:

"On the 1st of every month, analyze our website traffic data, 
create a report with key insights, and post it to our Notion database"

When to use: Monthly reports, billing cycles, or regular complex analysis.

When to Use Scheduled Tasks

Perfect Use Cases ✅

Recurring Research:

Daily news summaries on your industry
Weekly competitor product updates
Monthly market trend analysis
Real-time mentions and social listening

Regular Reports:

Daily sales metrics and pipeline updates
Weekly team performance summaries
Monthly analytics and KPI reports
Quarterly business reviews

Periodic Data Collection:

Daily price tracking for competitors
Weekly job posting monitoring
Monthly expense tracking and analysis
Continuous product review aggregation

Automated Monitoring:

Check for competitor updates weekly
Monitor market changes daily
Track industry news and announcements
Watch for regulatory or policy changes

Not Suitable For ❌

Immediate One-Off Tasks:

Don’t use Scheduled Tasks for tasks you need right now
Just run them manually instead

Real-Time Monitoring:

Scheduled Tasks run at specific times, not continuously
If you need constant monitoring, consider a different approach

Time-Sensitive Responses:

Tasks require time to execute (usually 5-30 minutes)
Not suitable for instant alerts or notifications

Real-World Scheduled Task Examples

Example 1: Daily News Digest

Purpose: Stay informed on industry news without manual research Task Name: “Daily AI News Summary” Configuration:

Schedule: Every weekday at 7 AM
Assigned Persona: “Industry Analyst”

Task Details:

"Research AI industry news from the past 24 hours. 
Focus on: 
- Funding announcements
- Product launches
- Major partnerships
- Regulatory developments

Deliver as:
- 5-bullet summary
- Email to my inbox
- Include source citations
- Keep to 2-minute read time"

Output Format:

Email subject: “[DATE] AI Industry News Summary”
5 key bullets with links
Include sources with publication names

Expected Result: Every weekday morning, a curated 5-bullet AI news summary in your inbox before work starts.

Example 2: Weekly Competitor Tracking

Purpose: Stay aware of competitor moves and market changes Task Name: “Weekly Competitor Intelligence” Configuration:

Schedule: Every Friday at 5 PM
Assigned Persona: “Competitive Intelligence Officer”

Task Details:

"Monitor these 10 competitors for updates:
[List competitors]

Check for:
- New blog posts and publications
- Product updates and feature releases
- Pricing changes
- Job postings (hiring/scaling signals)
- Press releases and announcements

Deliver as:
- Comparison table showing all changes
- Highlight significant moves
- Include URLs to source materials
- Add strategic implications

Save to:
- Google Drive shared folder
- Slack #competitive-intel channel"

Expected Result: Every Friday evening, a comprehensive competitive intelligence update covering all monitored competitors.

Example 3: Monthly Analytics Report

Purpose: Regular performance tracking and insights Task Name: “Monthly Website Analytics” Configuration:

Schedule: 1st of every month at 9 AM
Assigned Persona: “Data Analyst”

Task Details:

"Analyze website traffic from last month.

Data to Include:
- Monthly visitor trends (vs. previous month)
- Top 10 pages by traffic
- Traffic sources breakdown
- User behavior metrics
- Conversion funnel analysis
- Device/browser breakdown

Create a slide deck with:
- Executive summary (1 slide)
- Key metrics highlight (1 slide)
- Detailed analysis (3-5 slides)
- Trends and insights (2 slides)
- Recommendations (1 slide)

Post to: Slack #marketing channel
Include: Month-over-month comparisons"

Expected Result: On the 1st of each month, a comprehensive analytics dashboard with insights and recommendations posted to your marketing team.

How to Set Up Scheduled Tasks

Step 1: Describe the Task Clearly

Be specific about what you want the agent to do. Vague descriptions lead to vague results. ❌ Too Vague:

"Monitor competitors daily"

✅ Specific:

"Every weekday at 9 AM, check these 5 competitors 
for: product updates, pricing changes, job postings, 
and press releases. Create a comparison table 
and email me the results."

Step 2: Specify the Schedule

Define exactly when and how often the task should run: Daily Options:

“Every day at 9 AM”
“Every weekday at 8 AM”
“Every weekday and Saturday at 6 AM”

Weekly Options:

“Every Monday at 8 AM”
“Every Friday at 5 PM”
“Every Tuesday and Thursday at 2 PM”

Monthly Options:

“On the 1st of every month at 9 AM”
“On the 15th and 30th of every month”
“Last Friday of every month at 4 PM”

One-Time Options:

“Tomorrow at 3 PM”
“Next Wednesday at 10 AM”
“In 2 hours”

Step 3: Define the Output

Specify how you want the results delivered: Email Output:

"Email me a summary at [your-email@company.com]"

Slack Integration:

"Post results to #marketing channel"
"Send me a Slack DM with the summary"
"Post to multiple channels: #sales and #leadership"

Google Drive:

"Save as a spreadsheet to my 'Reports' folder"
"Create a presentation and save to shared team drive"
"Update existing document with latest data"

Notion Integration:

"Add results to my Notion database"
"Update the monthly analytics page"
"Create a new Notion page with findings"

Multiple Outputs:

"Email me a summary AND post to Slack #leadership"
"Create a report, save to Drive, and post to Notion"

Step 4: Assign the Right Persona

Connect the task to a custom Persona that has relevant expertise: Persona Assignment Examples:

Daily News Task → “Industry Analyst” Persona
Competitor Research → “Competitive Intelligence Officer” Persona
Analytics Report → “Data Analyst” Persona
Social Media Monitoring → “Social Media Manager” Persona
Content Curation → “Content Curator” Persona
Financial Analysis → “Financial Analyst” Persona

Managing Your Scheduled Tasks

View Active Schedules

Navigate to Settings → Scheduled Tasks to see all your active schedules: Dashboard View Shows:

✓ Task name and description
✓ Assigned Persona
✓ Schedule (when it runs)
✓ Output destination
✓ Last execution time
✓ Next scheduled execution
✓ Execution status (success/error)
✓ Task execution history

Pause a Schedule

Temporarily disable a schedule without deleting it: Use Cases:

During vacation (pause daily reports)
Holiday periods (stop competitor monitoring)
Temporary pause (resume after project)
Seasonal tasks (pause in off-season)

Edit a Schedule

Modify any aspect of an active schedule: What You Can Change:

✓ Schedule timing (frequency, time of day)
✓ Task description or requirements
✓ Output method or destination
✓ Assigned Persona
✓ Data sources or parameters

Delete a Schedule

Remove schedules you no longer need: Before Deleting:

⚠️ Confirm you don’t need historical data
⚠️ Check if downstream processes depend on it
⚠️ Consider pausing instead of deleting

View Execution History

See past runs, results, and any errors: Execution History Shows:

✓ Date and time of execution
✓ Task completion status (success/failed)
✓ Output delivered (email/Slack/Drive/etc.)
✓ Execution duration
✓ Any error messages
✓ Download past results
✓ Re-run a specific execution

Best Practices for Scheduled Tasks

1. Be Specific About Timing ✅

❌ Vague:

"In the morning"
"When you get a chance"

✅ Specific:

"Every weekday at 8 AM EST"
"Every Monday at 9:00 AM Pacific"
"On the 1st of the month at 6 AM UTC"

Why it matters: Vague timing leads to inconsistent execution. Specific times ensure predictable, reliable automation.

2. Define Clear Outputs ✅

❌ Unclear:

"Let me know what you find"
"Send me something about competitors"

✅ Clear:

"Email me a 5-bullet summary to [email@company.com]"
"Post a comparison table to Slack #competitive-intel"
"Create a slide deck and save to Google Drive /Reports folder"

Why it matters: Clear output specifications ensure you get results in the format you need.

3. Include Time Ranges for Research ✅

❌ Ambiguous:

"Research recent news"
"Find new information"

✅ Specific:

"News from the past 24 hours only"
"Updates since last week's report"
"New developments from last 30 days"
"Products released in the past month"

Why it matters: Time ranges prevent duplicate research and ensure fresh, relevant data.

4. Test Before Scheduling ✅

Testing Process:

Step 1: Run the task manually once
- → Ask Neuro to execute the task now
Step 2: Review the output
- → Check format, quality, and completeness
Step 3: Make adjustments
- → Request changes to the description
Step 4: Then schedule it
- → Once you’re happy, set up the schedule

Why it matters: Testing prevents surprises when the task runs automatically.

5. Set Realistic Time Expectations ✅

Typical Execution Times: Simple tasks: 3-5 minutes

News summary
Price tracking
Basic research

Medium tasks: 5-15 minutes

Competitor analysis
Market research
Report compilation

Complex tasks: 15-30 minutes

Deep analysis
Multiple data sources
Comprehensive reports

Scheduling tip: Schedule complex tasks at times when you don’t need immediate results.

Using Agents: Practical Workflows

Single Agent Workflow

Request:

"@ExecutiveSummarizer summarize this 50-page 
research report for our CEO"

Agent Process:

Reads entire report
Identifies key insights
Extracts critical metrics
Structures as executive summary
Delivers 1-page summary

Output:

Situation (2-3 sentences)
Key metrics (3-5 numbers)
Issues/Opportunities (Top 3)
Recommendations (1-3 actions)
Next steps (Clear timeline)

Multi-Agent Workflow (Coming Soon)

Request:

"Create a complete investor pitch deck"

Agent Collaboration:

@MarketResearcher
- → Analyzes market size and trends
- → Delivers market research data
@CompetitiveAnalyst
- → Researches competitors
- → Delivers competitive analysis
@FinancialAnalyst
- → Creates financial projections
- → Delivers financial models
@PitchDeckDesigner
- → Synthesizes all inputs
- → Creates professional deck

Output: Complete investor-ready pitch deck with market analysis, competitive positioning, and financial projections

Sequential Agent Workflow

Content Production Pipeline: Step 1: @ResearchAgent

"Research AI automation trends for 2025"
→ Output: Research report with insights

Step 2: @ContentWriter

"Using this research, write a blog post"
→ Output: 1500-word article

Step 3: @Designer

"Create visuals for this blog post"
→ Output: Graphics and charts

Step 4: @SocialMediaManager

"Create social posts promoting this article"
→ Output: 5 social media posts

Step 5: @EmailMarketer

"Create newsletter featuring this content"
→ Output: Email campaign

Result: Complete content pipeline from research to multi-channel distribution

Managing Your Agent Team

Agent Library Organization

Organize by Function:

Strategic Planning:
├─ Strategic Analyst
├─ Market Researcher
├─ Competitive Intelligence Officer

Operational:
├─ Project Manager
├─ Operations Analyst
├─ Process Optimizer

Sales & Marketing:
├─ Sales Strategist
├─ Content Creator
├─ Customer Researcher

Finance:
├─ Financial Analyst
├─ CFO Advisor
├─ Investor Relations Specialist

Product:
├─ Product Manager
├─ UX Researcher
├─ Feature Architect

Switching Between Agents

Use different agents for different contexts: Monday Morning: Strategy Meeting

→ @StrategicAnalyst creates market analysis

Tuesday: Sales Prep

→ @SalesStrategist creates battle cards

Wednesday: Investor Meeting

→ @InvestorRelations creates pitch deck

Thursday: Board Meeting

→ @ExecutiveSummarizer summarizes reports

Agent Handoff

When tasks need multiple perspectives:

Start with @MarketResearcher

"Research AI safety market for 2025"
→ Market research output

Pass to @ProductStrategist

"Based on this research, create roadmap"
→ Product roadmap output

Pass to @InvestorRelations

"Create pitch deck from roadmap and research"
→ Investor presentation output

Result: Fully integrated strategic materials from collaborative agent workflow

Best Practices for Agents

Creating Custom Agents

Do’s ✅

✓ Be specific about role and function
✓ Define clear expertise areas
✓ Include example outputs
✓ Test before deploying
✓ Iterate based on results

Don’ts ❌

✗ Create overly broad agents
✗ Combine conflicting responsibilities
✗ Skip testing phase
✗ Ignore user feedback
✗ Make agents too generic

Configuring Personas

Do’s ✅

✓ Define expertise clearly
✓ Specify communication style
✓ Include quality standards
✓ Provide example formats
✓ Document work preferences

Don’ts ❌

✗ Be vague about requirements
✗ Create conflicting instructions
✗ Overload with too many rules
✗ Skip quality definitions
✗ Forget to update regularly

Using Agents Effectively

Do’s ✅

✓ Match agent to task type
✓ Use multi-agent collaboration
✓ Provide clear instructions
✓ Review and refine outputs
✓ Give feedback for improvement

Don’ts ❌

✗ Use wrong agent for task
✗ Expect perfection first try
✗ Skip reviewing outputs
✗ Ignore quality issues
✗ Forget to iterate

Troubleshooting Agents

Agent Quality Issues

Problem: Inconsistent output quality Causes:

Vague persona instructions
Agent is too broad
Missing quality standards

Solutions:

Make persona more specific
Add example outputs
Define clear quality criteria
Test and iterate

Problem: Agent doesn’t follow style Causes:

Style preferences not explicit
Conflicting instructions
Missing examples

Solutions:

Add detailed style guide
Provide example outputs
Remove conflicts
Include brand guidelines

Problem: Agent misses domain knowledge Causes:

Expertise not documented
Knowledge Base incomplete
No domain training

Solutions:

Document domain expertise
Add to Knowledge Base
Include reference materials
Provide training examples

Getting Started with Agents

Quick Start (15 Minutes)

Step 1: Create First Agent (5 min)

→ Click “Create New” in Agent Store
→ Describe: “Create a market research agent”
→ Test in preview
→ Save agent

Step 2: Create First Persona (5 min)

→ Define expertise areas
→ Set communication style
→ Add output standards
→ Save configuration

Step 3: Combine and Test (5 min)

→ Assign persona to agent
→ Test with real task
→ Review output quality
→ Adjust as needed

Result: First specialized agent ready to use

First Week Plan

Day 1: Create 3 Core Agents

☐ Create sales/marketing agent
☐ Create research/analysis agent
☐ Create content/writing agent

Day 2-3: Build Personas

☐ Define 2-3 personas
☐ Document expertise and style
☐ Add quality standards

Day 4-5: Integration & Testing

☐ Assign personas to agents
☐ Test with real tasks
☐ Refine based on outputs

Day 6-7: Advanced Usage

☐ Try multi-agent workflows
☐ Create agent library
☐ Share with team

Result: Functioning agent team in one week

Measuring Agent Success

Key Metrics

Efficiency Metrics:

→ Time saved per task
→ Reduction in manual work
→ Faster completion times

Quality Metrics:

→ Output accuracy
→ Consistency level
→ Revision requirements

Adoption Metrics:

→ Agent usage frequency
→ Team adoption rate
→ Workflow integration

Success Indicators

Agents are working when:

✓ Outputs require minimal editing
✓ Quality is consistent
✓ Team uses regularly
✓ Workflows are faster
✓ Results meet standards

Conclusion

NeuroAI’s complete Agent system transforms how you work with AI. Instead of generic assistants, you build a specialized AI workforce tailored to your needs through five powerful layers: Layer 1: Custom Agents — Define WHAT your AI workers do Layer 2: Personas — Define HOW they work Layer 3: Tools — Enable WHICH capabilities they use Layer 4: Apps — Connect WHERE they access data Layer 5: Knowledge Base — Provide WHAT they know The result: An AI team that understands your business, follows your processes, uses the right tools, accesses your data, and delivers consistent, quality results—every time.

Last updated: January 2026

First Steps

Features

​The Complete Agent System

​Layer 1: Custom Agents (WHAT to do)

​Layer 2: Personas (HOW to work)

​Layer 3: Tools (WHICH capabilities to use)

​Layer 4: Apps (WHERE to access data)

​Layer 5: Knowledge Base (WHAT to know)

​Layer 1: Custom Agents

​What is a Custom Agent?

​Key Features

​One-Prompt Creation

​Reusability

​Collaboration (Coming Soon)

​Types of Custom Agents

​1. Role-Based Agents

​2. Industry-Specific Agents

​3. Task-Specific Agents

​Creating Custom Agents

​Creation Process

​Using Custom Agents

​Invocation Methods

​Managing Custom Agents

​Custom Agent Store

​Layer 2: Personas

​What is a Persona?

​Why Use Personas?

​The Problem with Generic AI

​The Persona Solution

​Persona Components

​1. Core Purpose

​2. Expertise Areas

​3. Communication Style

​4. Output Format

​5. Work Preferences

​6. Quality Standards

​Creating Personas

​Step 1: Define Purpose

​Step 2: Define Expertise

​Step 3: Set Communication Style

​Step 4: Define Work Preferences

​Step 5: Add Constraints

​Step 6: Define Success Metrics

​Persona Examples

​Example 1: Executive Strategist

​Example 2: Startup Fundraiser

​Example 3: Healthcare Strategist

​Combining Custom Agents + Personas

​The Power of Integration

​Real-World Integration Example

​Layer 3: Tool Configuration

​What is Tool Configuration?

​Available Tools (16 Categories)

​1. Task Management

​2. Files & Folders

​3. Web Search & Research

​4. File Reader & Knowledge Base

​5. Web Browser

​6. Presentations

​7. Image Processing

​8. Automation & Triggers

​9. Credentials & Authentication

​10. Research Tools

​11. Voice Calls

​12. Knowledge Base

​13. Agent Configuration

​14. Chat & Messages

​Tool Configuration Best Practices

​✅ Do’s

​❌ Don’ts

​Common Configuration Scenarios

​Scenario 1: Research Agent

​Scenario 2: Content Creation Agent

​Scenario 3: Automation Agent

​Scenario 4: Business Research Agent

​Layer 4: Apps Integration

​What are Apps?

​How Apps Work With Custom Agents

​Agent Capabilities With Apps

​Agent Personas & Their Apps

The Complete Agent System

Layer 1: Custom Agents (WHAT to do)

Layer 2: Personas (HOW to work)

Layer 3: Tools (WHICH capabilities to use)

Layer 4: Apps (WHERE to access data)

Layer 5: Knowledge Base (WHAT to know)

Layer 1: Custom Agents

What is a Custom Agent?

Key Features

One-Prompt Creation

Reusability

Collaboration (Coming Soon)

Types of Custom Agents

1. Role-Based Agents

2. Industry-Specific Agents

3. Task-Specific Agents

Creating Custom Agents

Creation Process

Using Custom Agents

Invocation Methods

Managing Custom Agents

Custom Agent Store

Layer 2: Personas

What is a Persona?

Why Use Personas?

The Problem with Generic AI

The Persona Solution

Persona Components

1. Core Purpose

2. Expertise Areas

3. Communication Style

4. Output Format

5. Work Preferences

6. Quality Standards

Creating Personas

Step 1: Define Purpose

Step 2: Define Expertise

Step 3: Set Communication Style

Step 4: Define Work Preferences

Step 5: Add Constraints

Step 6: Define Success Metrics

Persona Examples

Example 1: Executive Strategist

Example 2: Startup Fundraiser

Example 3: Healthcare Strategist

Combining Custom Agents + Personas

The Power of Integration

Real-World Integration Example

Layer 3: Tool Configuration

What is Tool Configuration?

Available Tools (16 Categories)

1. Task Management

2. Files & Folders

3. Web Search & Research

4. File Reader & Knowledge Base

5. Web Browser

6. Presentations

7. Image Processing

8. Automation & Triggers

9. Credentials & Authentication

10. Research Tools

11. Voice Calls

12. Knowledge Base

13. Agent Configuration

14. Chat & Messages

Tool Configuration Best Practices

✅ Do’s

❌ Don’ts

Common Configuration Scenarios

Scenario 1: Research Agent

Scenario 2: Content Creation Agent

Scenario 3: Automation Agent

Scenario 4: Business Research Agent

Layer 4: Apps Integration

What are Apps?

How Apps Work With Custom Agents

Agent Capabilities With Apps

Agent Personas & Their Apps

1. Sales Agent

2. Customer Success Agent