Adhere to the following guidelines when identifying the business use case:

1. Clearly define the specific problem or process you wish to automate with the AI agent.

2. Use tools such as Visio, Miro, and other similar tools to graphically lay out the problem or process you wish to automate.

3. Assess the potential impact and benefits of automating this use case, such as improved efficiency, reduced costs, or enhanced customer experience.

4. Identify the key KPIs you're going to measure to determine the ROI and prove the value.

• Start by defining a goal—Clearly articulate the primary objective of the AI agent, such as resolving customer inquiries or processing orders efficiently.

• Define the journey—Clearly identify the questions, actions, and features you want your AI Agent to have.

• Add knowledge—Integrate relevant knowledge bases the agent can access to provide accurate information.

• Define actions—Specify the actions the agent needs to perform and integrate necessary APIs or function calls.

• Preview—Preview your AI agent with knowledge and actions.

• Add instructions—Provide detailed instructions to enhance the accuracy and reliability of the agent's responses.

• Test and validate—Use platform preview tools to test the AI agent's performance and make necessary adjustments.

This section outlines best practices for writing goal prompts for the autonomous AI agent and actions to fulfill user intents.

Do's

• Keep the goal short, concise, general, and broad.

• Focus on the overall function or purpose of the bot.

• Consider the end result or benefit for the user.

• Use clear and concise language.

• Ensure the goal aligns with the actions and capabilities of the bot.

Don'ts

• Don't include specific details like locations, dates, or user information.

• Avoid mentioning particular actions or implementation methods.

• Don't use technical jargon or complex terminology.

• Avoid overly long or complicated goal statements.

• Don't include multiple unrelated goals in a single prompt.

• Avoid using ambiguous or vague language.

Adhere to the following recommendations while creating and managing knowledge base:

• Organize content logically. Make use of categories when creating your own knowledge document in the AI Agent studio.

• When uploading files, avoid any conflicting or duplicate information across documents.

• Check document quality before uploading.

• Split large files into smaller files if needed.

• Periodically review the knowledge and update whenever required.

Adhere to the following recommendations when creating actions:

• Clearly define action objectives in action description.

• Minimize complexity, keep actions simple.

• Accurately describe each entity/slot as this improves the accuracy of the LLM to better understand the task.

• Do not create conflicting or contradictory actions.

• Create deterministic logic in Connect flow for higher accuracy instead of relying on LLM.

Refer to the following tips when writing instructions for your autonomous AI agents:

• Keep it simple—Use clear, concise language. Avoid technical jargon or overly complex sentences.

• Use markdown—Use headings and ordered/unordered list markdown for best results.

• State your AI agent's identity—Begin by clearly defining the agent's persona (e.g., "You're a helpful customer support agent...").

• Break it down—Outline tasks step by step. For instance, "First, confirm your account number. Then, describe your issue."

• Plan for errors—Include fallback phrases such as, "I'm sorry, could you please repeat that?" if the input isn't clear.

• Preserve context—Remind the agent to remember previous responses to ensure continuity in long conversations.

• Reference actions—Clearly instruct how to use external actions at different steps. Make sure the referenced actions are enabled in the Actions tab to avoid any unexpected behavior.

• Add guardrails—Instruct the AI Agent to respond only in the context of the goal.

• Add examples—To improve accuracy, add examples wherever needed.