CPQ reasoning at DTU

AI-assisted CPQ training - from theory to working outcome

Here is some information about the upcoming training session at DTU on March 24–25.

This training is designed for CPQ, product, and sales leaders who want to explore how AI can be applied to configuration without compromising correctness, governance, or existing investments.

Rather than focusing on tools or future promises, the training introduces a practical operating model - how to augment your current CPQ landscape with AI-based reasoning by clearly separating context for decision-making from rules for correctness. This distinction allows AI to add speed and usability, while existing CPQ logic continues to guarantee valid configurations and pricing.

The format is deliberately compact - two half-day sessions - to ensure focus and momentum. Participants work hands-on with a reference example first, then apply the method to a real slice of their own product. The objective is not a conceptual demo, but a concrete outcome that can be reviewed and evaluated internally.

What this training is:

• A structured introduction to AI-assisted CPQ reasoning
• A hands-on exercise in onboarding and validating product logic
• A pragmatic way to explore AI augmentation without replacing existing systems

What this training is not:

• A CPQ replacement initiative
• A generic AI or prompt-writing course
• A theoretical program without tangible results

By the end of the second session, finishing by lunch on Day 2, participants will have a shared understanding of how AI can be introduced responsibly into CPQ, along with a working and explainable example grounded in their own product reality.

Training agenda (2 half-days)

The training is designed as two focused half-day sessions: afternoon on Day 1 and morning on Day 2. We finish by lunch on Day 2. The format is hands-on throughout, with short modules and built-in breaks to keep momentum.

Day 1 — Afternoon (12:00–17:00)

• 12:00–12:30 | Sandwiches (optional)
• 12:30–12:50 | Introduction
  • Why AI and reasoning matter for CPQ
  • Tools, working model, and goals for the two sessions
• 12:50–13:20 | Module 2: Truck example (reference case)
  • What an LLM can and cannot do as a configurator
  • Why constraints are essential for correctness and governance
• 13:20–13:30 | Break
• 13:30–13:50 | Roundtable: Business cases for AI in CPQ
  • Map your priorities: speed, correctness, sales enablement, onboarding, support
  • Define a realistic “slice” to build on Day 2
• 13:50–14:20 | Module 3: Onboarding a new product
  • Hands-on onboarding exercise and prompt modelling
  • Writing effective narratives; introducing modules and variants
• 14:20–14:30 | Break
• 14:30–15:10 | Module 4: Tuning with PRAG
  • Refining logic and narratives hands-on
  • Testing the assistant conversation and improving outputs
• 15:10–15:20 | Break
• 15:20–16:00 | Module 5: End-to-end testing (proposal flow)
  • Scenario walkthroughs and discussion
  • Identify gaps and define what “correct and explainable” means for your team
• 16:00–16:30 | Module 6: Wrap-up & reflection
  • Key takeaways, decisions, and next steps for Day 2 build
• 16:30–17:00 | Short introduction & challenge briefing
  • Recap of the method: Context vs Rules
  • Explanation of the challenge, success criteria, and available prepared data
  • Teams of two are formed

Day 2 — Morning (challenge-based, 09:00–12:30)

• 09:00–11:00 | Team challenge: build and validate
  • Each team works hands-on with a prepared product dataset
  • Define reasoning context (trade-offs, scenarios, explanations)
  • Implement a small set of strict rules for correctness
  • Test configurations end-to-end to ensure explainability and valid outcomes
• 11:00–11:10 | Short break
• 11:10–11:50 | Team presentations
  • Each team presents their solution to the group
  • Walkthrough of reasoning, rules, and resulting configurations
  • Group discussion on differences, insights, and modeling choices
• 11:50–12:00 | Wrap-up & close
  • Key takeaways from the challenge
  • Discussion on how the approach translates to real CPQ environments
  • Next steps and follow-up sessions
• 12:00–12:30 | Sandwiches