Week 1: Operations Audit

Finding where AI actually helps

Today’s topics

  • How established companies differ from startups
  • Building business cases that work
  • Mapping your current operations
  • Spotting worthwhile opportunities

The established business difference

  • Legacy infrastructure serving real customers
  • Teams with established workflows
  • Service commitments requiring consistency
  • Business data trapped in systems

The expensive mistake

Trying to automate every process you find.

AI projects fail when companies pick wrong work to automate.

Figure out what is worth automating first.

Where AI creates value

  • Repetitive work requiring judgment
  • Pattern detection across large datasets
  • High-volume language tasks
  • Consistency in customer communications

Where AI wastes money

  • Work requiring extensive business context
  • Highly variable processes
  • Relationship-dependent work
  • Decisions with compliance risk

Calculate first

  • What does this cost today?
  • How much improvement is realistic?
  • What will implementation cost?
  • What does maintenance run?

Map before automating

Every process has three components:

  • Inputs: What triggers the work
  • Processing: Steps to complete it
  • Outputs: What gets produced

Where time goes

  • Direct work time
  • Waiting periods
  • Rework cycles
  • Coordination overhead

Quality problems AI solves

  • Inconsistent customer responses
  • Data entry mistakes
  • Forgotten follow-up tasks
  • Knowledge trapped in people’s heads

Integration difficulty

Rate system connection complexity:

  • Simple: Standalone work
  • Moderate: One or two systems
  • Complex: Three or more systems

The opportunity matrix

Plot processes on two dimensions:

  • Horizontal: Value if improved
  • Vertical: Difficulty to implement

Quick wins

  • Drafting email responses
  • Generating weekly reports
  • Summarizing documents
  • Cleaning data before entry

Strategic plays

  • Intelligent customer service
  • Document processing
  • Predictive analytics
  • Quality control automation

Workshop: Process inventory

  1. List 20-30 major processes
  2. Note biggest pain point for each
  3. Estimate weekly time and monthly cost
  4. Rate AI suitability (1-5 scale)

Workshop: Top opportunities

For your five highest-rated processes:

  • Current state metrics and costs
  • Desired future state with numbers
  • Integration needs
  • Resource requirements

Key takeaway

Pick the right opportunities, not the most advanced technology.

Audit thoroughly. Prioritize on clear value. Build from wins.