AI in Requirements Engineering

This dimension assesses how AI assists in capturing, refining, and prioritising requirements.

Sample assessment questions for each level:

  • Level -1: “Is the use of AI tools for requirements analysis explicitly prohibited or discouraged?”
  • Level 0: “Are AI requirements tools used inconsistently by individual team members without standards?”
  • Level 1: “Has the organisation identified specific use cases for AI in requirements gathering?”
  • Level 2: “Are large language models used to draft requirement documentation?”
  • Level 3: “Is AI used to analyse customer feedback or behaviour to derive requirements?”
  • Level 4: “Does AI detect ambiguity or contradictions in requirements?”
  • Level 5: “Does AI proactively suggest backlog grooming or feature reprioritisation?”

Key metrics to track:

  • Requirements quality improvement: Percentage reduction in ambiguities after AI analysis
  • Customer feedback processing efficiency: Time saved using AI for analysis vs. manual methods
  • Predictive accuracy: Percentage of AI-predicted user needs that become actual requirements
  • Requirements clarity score: Measured through structured evaluation of AI-refined requirements
  • Feature prioritisation alignment: Correlation between AI-suggested priorities and actual user value