Klaus Breyer Tech Leadership, Product Delivery & Startup Strategy.

Shape Up for AI Products: Prototyping while Shaping

In software, outcomes can never be guaranteed. This is why Shape Up excels with its fixed-time, variable-scope approach, allowing strategic deployment of ressources.

But, outcomes can be even harder to guarantee when you build products on top of some form of AI like LLMs. This is why de-risking projects with a structured yet flexible approach like Shape Up is even more important.

Let me explain.

Framing - Laying the Foundation

The journey begins with Framing. This initial phase is all about understanding and articulating:

  1. The Problem: What specific challenge or need is the AI project addressing?
  2. The Context: What are the environmental, technological, and business contexts in which this problem exists?
  3. The Desired Outcome: What does success look like for this project? What impact or results are expected?

Framing sets the stage, offering a clear, focused lens through which the entire project is viewed. It helps prevent mission creep and keeps the project aligned with its core objectives.

Shaping - Crafting the Solution Space

Once the frame is set, Shaping takes the helm. This phase involves:

  1. Exploring Solutions: Brainstorming potential ways to address the framed problem.
  2. Refining Ideas: Narrowing down the ideas to those most viable and aligned with the project’s goals.
  3. Prototyping: Testing and experimenting with these ideas to see what works and what doesn’t in a low-risk environment.

Shaping is where creativity meets pragmatism. It’s a time for open-ended exploration, but always within the boundaries set by the Framing phase.

Prototyping: Bridging Shaping and Building

By weaving Prototyping into the Shaping phase, Shape Up creates a dynamic and responsive environment for AI project development. It allows teams to:

  1. Test Feasibility: Determine if the ideas generated during Shaping are technically and practically feasible.
  2. Gather Early Feedback: Obtain insights from stakeholders, users, or team members, refining the approach based on this feedback.
  3. Risk Mitigation: Turn abstract concepts into tangible prototypes, reducing the unknowns and uncertainties inherent in AI development.

Prototyping is not about creating a polished product; it’s about learning, experimenting, and adapting.

Conclusion: A Strategic Approach for AI Projects

That engineers need to prototype to build an AI product is clear. But you don’t need to pause your product process while they are doing it.

Prototyping while Shaping makes both Shaping and the Prototyping of the tech more efficient.