Module 4: Impact Assessment — Where to Automate First

Duration: 60 minutes Day: Day 1, Session 4

What This Module Is About

Not everything should be automated, and not everything can be automated well with today's AI. This module gives you a structured framework for evaluating your operational tasks and identifying which ones will deliver the highest return from an AI agent.

The output of this module is your capstone candidate — the operational problem you'll build a domain agent for on Day 3.

Learning Objectives

By the end of this module, you will be able to:

1. Score your top 10 operational tasks on the Automation Quadrant using frequency × complexity analysis
2. Apply the Automation Quadrant to identify candidates in each quadrant: automate now, automate with care, augment, leave alone
3. Distinguish between tasks that benefit from context engineering alone vs. tasks that require a tool-enabled agent
4. Select your capstone candidate — the operational task you'll build a domain agent for on Day 3
5. Draft a one-paragraph problem statement for your capstone that captures the task, its current manual cost, and the expected agent behavior

Solo Learner

This exercise is designed for solo completion. The scoring template works whether you are in a live workshop team or learning independently on Udemy. If you do not have a live workshop group, you complete the scoring, table share, and problem statement steps on your own — the output (your capstone candidate and problem statement) is identical.

Prerequisites

• Modules 1-3 complete
• Bring a list of 10 operational tasks from your own environment (or use the provided examples)

Module Contents

Section	Content	Time
Lab	Automation Quadrant Scoring Template	40 min
Reading	Evaluating Automation Candidates	15 min
Quiz	Module 4 Assessment	5 min
Exploratory	Stretch Projects	Optional

Key Concept: The Automation Quadrant

High Frequency
      │
  A   │   B
Augment│Automate
      │  Now
──────┼──────── High Complexity
      │
  D   │   C
Leave │Automate
Alone │with Care
      │
Low Frequency

Quadrant B (High Frequency + Low Complexity): Automate now — these are your highest-ROI targets. Daily alert triage, routine health checks, standard deployments.

Quadrant A (High Frequency + High Complexity): Augment — AI assists the human, doesn't replace them. Complex incident diagnosis, architectural decisions, novel failure modes.

Quadrant C (Low Frequency + High Complexity): Automate with care — occasional but high-stakes tasks. Disaster recovery, major version upgrades.

Quadrant D (Low Frequency + Low Complexity): Leave alone — the manual effort is so low that automation overhead isn't worth it.

Your capstone candidate should land in Quadrant B or the high end of Quadrant A.