AI Assistant Workforce
An AI Assistant Workforce is when multiple AI Assistants work together to complete complex tasks that span across different platforms and tools. Instead of using a single Assistant, you orchestrate several Assistants to collaborate on a unified goal.See an Assistant Workforce in Action
Watch a real-world example of an AI Assistant Workforce in our Relevance AI communityClick here to view the workforce example →
AI Assistant Workforce vs. Regular Workforce
AI Assistant Workforce
Uses pre-built AI Assistants from the Marketplace designed for specific platforms
- Ready-to-use Assistants (Notion, Gmail, HubSpot, etc.)
- No setup required beyond connecting integrations
- Optimized for common platform tasks
- Best for cross-platform workflows
Regular Workforce
Uses custom Agents you build yourself for specific business needs
- Tailored to your unique processes
- Can include custom tools and logic
- Requires Agent building and configuration
- Best for specialized workflows
Can you mix and match?Yes! You can combine AI Assistants with your custom Agents in the same workforce:
- Use both pre-built Assistants and custom Agents together
- Create hybrid workflows that leverage the best of both
- Example: Use a custom research Agent to gather data, then have the Gmail Assistant send results to your team
How It Works
Each Assistant in your workforce specializes in one platform (Notion, Gmail, Google Docs, etc.) and has access to that platform’s full API capabilities. When you give a complex task that requires multiple platforms, the Assistants automatically:- Break down the task into platform-specific subtasks
- Execute in sequence - each Assistant completes its part
- Pass information between Assistants as needed
- Deliver the complete result - all steps handled automatically
The Technology Behind It
Flexible API Tools
Each Assistant uses adaptable tools that work across different tasks and platforms
Comprehensive Documentation
Each Assistant has full API documentation to understand their platform’s capabilities
Advanced AI Models
Uses models like GPT-4 to reason through tasks and make intelligent decisions
Self-Healing
Learns from mistakes and automatically retries operations for better results
What this means for you:
- Assistants can handle tasks they weren’t explicitly programmed for
- New capabilities emerge as APIs are updated
- Complex workflows work automatically without custom integration code
- Performance is optimized for efficiency
Workforce Example
Here’s an example of multiple Assistants and Agents working together:The Task
What Happens Behind the Scenes
This single request automatically coordinates a mix of Assistants and Agents:This workforce combines 3 pre-built Assistants (Google Slides, Google Drive, Gmail) with 1 research Agent to complete the task.
1
Research Agent (Agent)
Searches for relevant information about dinosaurs from knowledge sources or the web
2
Google Slides Assistant (Assistant)
Creates a presentation with the researched content, formats slides, adds structure
3
Google Drive Assistant (Assistant)
Creates the “DinosaurFacts” folder and saves the presentation to it
4
Gmail Assistant (Assistant)
Composes and sends an email with the presentation link to the specified recipient
Benefits of AI Assistant Workforces
Connect Any Platform
Work across tools that don’t normally integrate together
Handle Complex Tasks
Complete sophisticated workflows automatically
Use Plain English
No technical skills required—just describe what you want
Getting Started with Workforces
1
Get Your Assistants and Agents
For Assistants: Visit the Marketplace and clone each Assistant your workflow requiresFor Agents: Build custom Agents or clone existing ones for specialized tasks
2
Configure Everything
Assistants: Follow the Cloning and Setting up Assistant guide to set up API keys or direct integrationsAgents: Configure your custom Agents with the necessary tools and prompts
3
Create Your Workforce
In Relevance AI, create a workforce and add your configured Assistants and/or Agents to it
4
Give It a Task
Describe your complete workflow in natural language - the workforce will coordinate all Assistants and Agents automatically
Common Workforce Patterns
Content Creation Pipeline
Google Docs, Google Slides, Gmail, Notion working together to create reports, presentations, and distribute to teams
CRM to Communication
HubSpot, Gmail, Google Docs coordinating to manage leads and send personalized communications
Cross-Platform Data Migration
Multiple Assistants syncing data between platforms and providing migration summaries
Workflow Automation
Any combination of Assistants automating recurring tasks across your tools
Tips for Workforce Tasks
Be Explicit About the Workflow
Be Explicit About the Workflow
Mention each platform and what should happen in each step to ensure the workforce understands the complete task.Example: “Get my recent contacts from HubSpot, create a summary document in Google Docs, and email it to my team”
Specify the Order
Specify the Order
Use sequential language like “first”, “then”, “finally” when the order of operations matters for your workflow.Example: “First create the Notion database, then add sample projects, finally send me a summary email”
Include Identifiers
Include Identifiers
Provide specific file names, email addresses, folder names, or collection names so Assistants know exactly where to work.Example: “Save to folder ‘Q1 Reports’” or “Email to team@company.com”
Test Individual Steps
Test Individual Steps
Before running a complex multi-Assistant workflow, test that each Assistant can handle its part independently. This helps identify issues early and ensures smoother execution.
Available Assistants for Your Workforce
Browse the Marketplace
Clone Assistants from the Marketplace and add them to your workforceExample search: “Notion Assistant” or “Gmail Assistant”

