Decision Guide

AI Implementation Decision Guide

A structured methodology to evaluate and select the right AI implementation type for your specific product needs.

Key Assessment Questions

Consider these questions in the context of your specific use case.

1

Who initiates the action?

Determine if your AI product should be user-initiated or system-initiated.

Options to Consider:

User explicitly

May be better suited for tools or copilots where user control is paramount.

System automatically

May be better suited for workflows and agents where automation is the priority.

2

How predictable is the process?

Assess the predictability of the tasks you want to automate.

Options to Consider:

Highly predictable

Well-suited for workflows with defined steps.

Requires adaptation

May require copilots or agents with more flexibility.

3

Where does the AI operate?

Consider where in the user workflow the AI appears.

Options to Consider:

On demand (user calls it)

Tools or copilots that users explicitly invoke.

In workflow (part of process)

Workflows or infrastructure integrated into existing processes.

Behind scenes (invisible)

Agents or infrastructure that work autonomously.

4

What's the level of autonomy needed?

Determine how much independence the AI should have.

Options to Consider:

Low (human verification)

Tools that make suggestions for human approval.

Medium (supervised)

Workflows or copilots with human oversight.

High (fully autonomous)

Agents or infrastructure that can operate independently.

AI Type Recommendations

Based on your answers to the assessment questions, here's when each AI type is most appropriate.

Tools

Best For:

  • User-initiated actions
  • Specific, focused tasks
  • Situations requiring high control
  • Quick implementation needs

Workflows

Best For:

  • Predictable multi-step processes
  • Defined business processes
  • Medium autonomy requirements
  • When connecting existing systems

Copilots

Best For:

  • Collaborative work
  • Complex knowledge tasks
  • Situations requiring iterative refinement
  • When human judgment remains essential

Agents

Best For:

  • Autonomous operation
  • Complex decision-making
  • Adaptable problem-solving
  • When high independence is acceptable

Infrastructure

Best For:

  • Foundation for other AI capabilities
  • Enterprise-wide deployment
  • Strategic long-term investments
  • When building a platform

Next Steps

Once you've identified the most appropriate AI implementation type, consider these next steps:

1

Review Detailed Type Guide

Explore our detailed guide for your selected AI type to understand implementation considerations.

View AI Types
2

Use Implementation Guide

Follow our step-by-step implementation guide to plan your AI product development.

View Implementation Guide
3

Explore Case Studies

Learn from organizations that have successfully implemented similar AI approaches.

View Case Studies