AI Goals (Requirements)

AI Goals are the core directives that drive your autonomous agents. In the system, these are represented as Requirements - high-level objectives that define what needs to be achieved, leaving the how to the agents’ strategic planning.

Unlike a simple task with a single action (e.g., “send an email”), a Requirement is a broader objective (e.g., “Increase leads by 20%”) that requires a series of autonomous actions, analysis, and iterations.

How it Works: Token Budget and Execution

The engine that powers these goals is the Token Budget. Every Requirement is assigned a specific budget in tokens, which represents the computational resources and effort agents are allowed to spend on achieving it.

Continuous Execution

Agents will continue to execute actions, analyze data, and refine their strategies for a Requirement until the assigned budget is exhausted. This allows for complex, multi-step problem solving that evolves over time.

• Budget Assignment:
  • Automatic: The system can automatically assign a budget based on the estimated complexity and priority of the goal.
  • Manual: You can explicitly set a token limit to control costs and focus resources on high-priority objectives.

Strategic Planning (CMO Agent)

The CMO (Chief Marketing Officer) agent acts as the strategic planner for your Requirements. It analyzes each goal to:

1. Deconstruct the Objective: Break down the high-level Requirement into smaller, actionable steps.
2. Allocate Budget: Distribute the token budget across different sub-tasks and channels to ensure balanced execution.
3. Create Instance Plans: Generate specific plans that other specialized agents (like Copywriters or Sales Agents) will execute.

Limitations

Understanding the boundaries of AI Goals is crucial for success:

• Budget Constraints: Agents stop working on a goal once the token budget is depleted, even if the goal is not fully achieved. Setting an appropriate budget is critical for complex tasks.
• Ambiguity: Vague requirements lead to inefficient token usage. Agents need clear, quantifiable targets to function effectively.
• Resource Access: Agents are limited by the tools and data access you provide. They cannot perform actions they have not been given permission or tools for.
• Context Window: Agents may not have the full historical context of your business unless it is explicitly provided in the knowledge base or requirement description.

Best Practices

To maximize the effectiveness of your AI Goals:

• Be Specific and Quantifiable: Use numbers, percentages, and dates (e.g., “Generate 100 new qualified leads by end of Q3” instead of “Get more leads”).
• Allocate Sufficient Budget: Ensure the token budget is realistic for the complexity of the goal. A complex goal with a tiny budget will likely fail or stop midway.
• Prioritize: Focus on 1-3 primary objectives at a time. Too many concurrent goals can dilute the effectiveness of your agents.
• Feedback Loop: Regularly review agent performance and adjust the budget or goal description based on the results.