What Are Assistants?
Assistants are designed for recurring, single-purpose tasks that you perform frequently. Think of them as specialized helpers that apply consistent instructions, knowledge, and context to your work.
How Assistants Work
Assistants leverage:
Custom Instructions: Define tone, style, or specific guidelines
File Attachments: Upload reference documents for context
Knowledge Access: Connect to your organization's knowledge
Web Access: Retrieve real-time information when needed
Typical Use Cases
Brand Voice Editing: Ensure all content matches your brand guidelines
Compliance Checking: Review documents against regulatory standards
Technical Documentation: Generate consistent technical writing
Translation & Localization: Maintain tone across languages
What Are Agents?
Agents are built for complex, multi-step workflows that require autonomous decision-making or structured task execution. They can handle end-to-end processes that involve multiple tools, integrations, and conditional logic.
How Agents Work
Agents offer two execution modes:
Agent Mode (Flexible)
The Agent autonomously decides which steps to take based on the task at hand. It dynamically chooses tools, retrieves information, and adapts its approach.
Workflow Mode (Structured)
You define a specific sequence of steps, conditions, and actions. The Agent follows this structured workflow consistently every time.
What Agents Can Do
Combine Tools & Integrations: Connect CRM, project management, databases, and more
Process External Signals: React to webhooks, API calls, or scheduled triggers
Execute Multi-Step Logic: Handle conditional branching and complex decision trees
Autonomous Operation: Run tasks independently without manual intervention
Typical Use Cases
Customer Onboarding: Collect information, create accounts, send welcome emails
Data Processing Pipelines: Extract, transform, and load data across systems
Automated Research: Gather information from multiple sources and compile reports
Issue Triage & Routing: Analyze support tickets and route to appropriate teams
Key Differences at a Glance
Feature | Assistants | Agents |
Task Complexity | Single, recurring tasks | Multi-step, complex workflows |
Execution Style | Direct response in chat | Autonomous or structured execution |
Best For | Consistency & efficiency | Automation & orchestration |
Tools & Resources | Knowledge, Web, Files | Knowledge, Web, Files, Integrations, External Tools |
Setup Complexity | Simple: instructions + context | Advanced: workflows + logic |
Sharing | Workspace or Organization-wide | Organization-wide (planned) |
Interaction | Used in chat conversations | Triggered by events or schedules |
Example Use Case | "Review this text for brand compliance" | "Onboard new customer end-to-end" |
When to Use Assistants
Choose Assistants when you need:
✅ Consistent execution of repetitive tasks
Example: Every blog post needs to follow the same brand voice guidelines.
✅ Quick access to specialized knowledge
Example: Answer customer questions using your product documentation.
✅ Human-in-the-loop workflows
Example: You review and approve content before it's finalized.
✅ Team collaboration with shared context
Example: Your entire marketing team uses the same brand voice assistant.
✅ Simple setup without technical complexity
Example: Non-technical users can create and share assistants easily.
[SCREENSHOT 3 NEEDED: Assistant in Chat]
Beschreibung: Screenshot des Chat Interface mit geöffnetem Assistant Selector Dropdown. Zeigt Liste verfügbarer Assistants. User: Sarah Müller, DACH Region Workspace.
When to Use Agents
Choose Agents when you need:
✅ End-to-end process automation
Example: Automatically process incoming leads from form submission to CRM entry.
✅ Multi-system orchestration
Example: Pull data from your CRM, enrich it with web research, and update your database.
✅ Conditional logic and decision-making
Example: Route support tickets based on urgency, category, and customer tier.
✅ Scheduled or event-triggered execution
Example: Run a daily report compilation every morning at 8 AM.
✅ Autonomous operation without manual input
Example: Monitor data feeds and alert teams when anomalies are detected.
[SCREENSHOT 4 NEEDED (OPTIONAL): Workflow Builder]
Beschreibung: Screenshot des Workflow Mode Builders für Agents (falls verfügbar), zeigt strukturierte Workflow-Steps mit Conditionals und Actions.
Can I Use Both Together?
Absolutely! Assistants and Agents complement each other:
Assistants handle the "thinking" tasks: Drafting, reviewing, analyzing content with human oversight
Agents handle the "doing" tasks: Automating processes, moving data, triggering actions
Example Combined Workflow
Agent monitors customer support inbox and categorizes tickets
Agent routes urgent tickets to a Slack channel
Human reviews the ticket and uses an Assistant to draft a response
Agent logs the resolution in the CRM and closes the ticket
This combination lets you automate the routine while keeping humans in control of the nuanced decisions.
Getting Started
Ready to Create Assistants?
Assistants are available now in Assistant Studio. You can create, customize, and share them across your workspace or organization.
👉 Learn more: Creating Your First Assistant [blocked] (Link to relevant Help Center article)
Interested in Agents?
Agent Studio is currently in development. Agents will enable powerful workflow automation and multi-system orchestration.
👉 Stay updated: Check the Product Roadmap [blocked] or contact your account manager for early access opportunities.
Need Help?
If you have questions about which tool is right for your use case, reach out to our support team or consult with your organization's admin.
Related Articles: