Milan Stancevic
Product Stance · Milan Stancevic

AI can build fast. Product judgment decides what deserves to be built.

I am Milan Stancevic, a product manager in Munich. I wrote Worth Building as the playbook I wanted beside me when one question was still unresolved: is this worth building? Product Stance is where I put that question to work, in the products I build and the books I write.

If this works, what changes for the customer or the business?

I test every AI idea with this question. When building gets cheap, the hard part is choosing the problem worth solving, the evidence to trust, and whether a result moved a real outcome.

One stack, each built on the last

Five pieces of work, starting with what is live today. Each one builds on what the others taught me.

  1. Chef Mealan Live pilot

    A meal planning app for nutrition coaches and their clients. It sizes each meal to the client's calorie and protein targets. Live since September 2026 in a closed pilot.

    Built with Claude Code. Evals for the AI parts, automated tests and end-to-end walkthroughs before every release.

    Visit chefmealan.com →
    Chef Mealan: a meal photo read into foods and scored against the client's plan
  2. Worth Building Playbook

    Product thinking for the AI and agentic era. Six connected product decisions, from where to focus to what changed in real use. Human, code, AI and agents may each carry part of the work; the decisions stay with people.

    Read the free book →
    Worth Building, a product playbook. Product thinking for the AI and agentic era
  3. PlateMate Prototype

    The prototype before Chef Mealan, built with the Worth Building method. I wrote its eval suite before the product.

    Capstone Honors, AI Product Management Certificate, Product Faculty.

    Open the prototype and its evals → PlateMate evals: six cases judged by a human against what the PRD predicted, all six pass
  4. Xnumia Agentic payments

    Stress-testing the Xnumia idea in my AI canvas, plus prototypes in AI Studio, won me the product lead role on the founding team of this agentic payments platform. I then defined the product and its documentation.

  5. Continuous PM OS AI canvas

    My own AI canvas for testing product ideas. It walks an idea through the stages it needs to survive: which problem, for which segment, then the solution.

    Built in Google AI Studio with Gemini.

How I decide what is worth building

Worth Building treats product work as a chain of decisions, in this order. Each one needs evidence and ends in a verdict: go on, change course, or go back a step.

  1. 1 · Direction

    Where do we play?

    Start with value. Align with the business on the market, the customers and the goal.

  2. 2 · Discovery

    What are we solving, for whom?

    Prove the opportunity: the real problem, the people who have it, and the value gap.

  3. 3 · Worth solving

    Is this problem worth solving?

    Is it big enough for them and for the business, and is the evidence strong enough to act on?

  4. 4 · Worth building

    Is this worth building?

    Only then the solution: if it works, what changes for the customer or the business?

AI changes the work between these decisions. The decisions stay with people: choosing the problem, judging the evidence and setting the quality bar.

Background

12+ years as Product Manager and Product Owner in B2B and B2B2C software in Germany and Serbia, from MVPs and pilots to one platform rolled out in 12 countries.

  • Real Garant (Zurich Group), Product Owner, vehicle warranty platform
  • bookingkit, Senior Product Manager, reseller marketplace and integrations
  • Commerce Connector and Cuponation, Product Manager roles
  • Planer.rs, my own event platform, acquired by Ringier Axel Springer Serbia

Books

I am the author of Worth Building and Worthy Numbers, and a co-author of Insights from Extraordinary Coaches. Worth Eating comes next.

All books at chefmealan.com

Certificates and learning

  • AI Product Management Certificate, Product Faculty, with Capstone Honors (2026)
  • AI Mastery, Mindvalley, 108 CPD credits (2026)
  • Building with AI Agents, Turing College, from November 2026
  • Dipl. Ing. in Industrial Engineering, University of Belgrade

Get in touch

I am open to Senior AI Product Manager roles in Munich or remote in Germany. The fastest way to reach me is a message on LinkedIn.

Message me on LinkedIn