MeshMesh
Core Concepts

Tasks

Learn how tasks work in MeshMesh Studio - from planning to completion, with AI-powered automation for your Salesforce workflows.

Tasks & Workflows

A task in MeshMesh Studio represents a goal or objective you want to accomplish. Describe what you want in natural language, and your AI assistant figures out the steps needed to make it happen.

Tasks automatically track progress, create artifacts, and save results for future reference.

Types of Tasks

Single-Step Tasks

Simple, focused tasks that accomplish one thing:

  • Show me my Marketing Cloud templates
  • Create a customer segment of high-value buyers
  • Generate a campaign performance report

Multi-Step Workflows

Complex tasks requiring multiple actions:

  • Create a welcome email series with 3 emails
  • Set up a customer journey from acquisition to retention
  • Integrate data from multiple sources and create analysis

Iterative Tasks

Tasks that evolve through conversation:

  • Start with broad exploration
  • Narrow based on findings
  • Iterate until desired outcome achieved

Task Components

Task Lifecycle

You describe what you want to accomplish

AI analyzes your request and plans the approach

You review the plan (if needed) before execution begins

AI performs the necessary steps automatically

Progress tracked in real-time as each step completes

Artifacts created to document important discoveries

Results presented with summaries and visualizations

Artifacts available for detailed exploration

Follow-up questions can be asked

Results saved automatically

Artifacts preserved for future reference

Shareable with team members

Working with Tasks

Create Your Task

Choose from three methods:

From Scratch: Click "New Task", name it, describe what you want, and start

From Template: Browse template library, select matching goal, customize, and start

From Previous Task: Open completed task, click "Create Similar Task", modify, and start

Organize by Status

Keep Tasks Organized

Use categories, priorities, and status filters to manage multiple concurrent tasks.

  • Active: Currently in progress
  • Waiting: Paused, waiting for input
  • Completed: Finished successfully
  • Archived: Moved to archive

Search and Filter

Find tasks by name, description, artifacts created, date, platforms used, or team members involved.

Best Practices

Writing Effective Task Descriptions

Be specific, provide context, and state constraints clearly.

Examples

Be Specific:

  • ❌ "Do something with emails"
  • ✅ "Create a welcome email series with 3 emails for new customers"

Provide Context:

  • ❌ "Make a campaign"
  • ✅ "Create a Black Friday campaign targeting customers who abandoned carts in the last 30 days"

State Constraints:

  • ❌ "Build a journey"
  • ✅ "Build a customer journey that sends no more than 2 emails per week and respects opt-out preferences"

Breaking Down Complex Tasks

For very complex goals, start with a high-level task, review what AI discovers, create separate tasks for each major component, and link related tasks together.

Task Performance

Simple Tasks

Complete in 2-5 minutes. Quick operations and single-platform queries.

Medium Tasks

Complete in 5-15 minutes. Multi-step workflows and moderate complexity.

Complex Tasks

May take 15-30 minutes. Advanced integrations and extensive operations.

Tasks automatically timeout after 30 minutes. Break very large tasks into smaller pieces and save progress incrementally.

Troubleshooting

Tips & Tricks

  • Leverage Task History: Review what worked well and create templates from successful patterns
  • Run Concurrent Tasks: Multiple tasks run independently with results available as each completes
  • Iterate Freely: Refine based on results and adjust approach as needed
  • Share with Team: Collaborate on tasks and build organizational knowledge

Next Steps