Lean Startup COO: Agentic Automation Strategy prompt

As a Chief Operating Officer for a lean startup, your primary goal is to achieve a 70% reduction in back-office management time by implementing autonomous AI agents. Design a comprehensive agentic processing strategy based on the following phases:

Phase 1: Process Identification
Analyze the provided '{CURRENT_DAILY_OPS}' and identify 5 highly repetitive, time-consuming processes currently handled by humans that are prime candidates for full automation by autonomous AI agents.

Output: A numbered list of these 5 processes. For each, briefly explain why it's suitable for AI automation (e.g., rule-based, high volume, low variability, structured data).

Phase 2: Hand-off Protocol Design
From your identified list, select the single highest-friction process (the one causing the most human bottleneck). For this process, design a detailed 'hand-off protocol' for an autonomous AI agent. This protocol must clearly define:
1. Trigger: The specific, quantifiable event that initiates the AI agent's action (e.g., 'New customer support ticket created in Zendesk with severity 'Medium' or higher').
2. Action Sequence: A step-by-step description of the AI agent's autonomous operations, including necessary data retrieval, processing steps, and decision points (e.g., 'Agent parses ticket for keywords and customer ID,' 'Agent queries CRM for customer history and recent interactions,' 'Agent drafts initial response based on FAQ database').
3. Escalation Rules: Clear, conditional 'IF/THEN' rules defining when and how the AI agent should escalate to a human, or attempt automatic resolution (e.g., 'IF sentiment analysis indicates extreme negative tone, THEN flag for human review and assign 'Urgent' priority; IF solution found in FAQ, THEN apply solution and mark ticket 'Pending Customer Confirmation').

Output: A structured breakdown of the hand-off protocol for the selected process, using distinct headings for Trigger, Action Sequence, and Escalation Rules.

Phase 3: AI Agent System Prompt
Develop a comprehensive 'System Prompt' that you would provide to an autonomous AI agent to execute the hand-off protocol from Phase 2 without constant human supervision. This prompt must clearly define:
* Agent Role & Goal: The agent's persona, purpose, and specific objective for the task.
* Constraints & Guardrails: Any limitations, forbidden actions, or safety protocols.
* Access & Tools: Implicitly or explicitly, what systems or data the agent can interact with.
* Decision-Making Logic: How the agent should prioritize, make choices, and handle ambiguities.
* Error Handling & Fallbacks: Detailed instructions for dealing with unexpected inputs, system failures, or unresolvable situations, directly linking to your escalation rules.
* Output & Reporting: How the agent should communicate its actions, results, and status (e.g., log entries, summary reports, status updates in a specific format).

Output: The complete, ready-to-use System Prompt enclosed within a markdown code block.