PETER KORACA
Product and tech leader helping organisations design and build AI-powered products – fast.
AI Product sprint
AI Product Sprint is the methodology I’ve refined over several years across real projects to help teams strategically design and build an AI product quickly – without falling into “design theatre”.
In 8 weeks, you’ll have:
A working alpha/MVP prototype (demo-ready, testable, not slideware)
A clear interaction & governance model (what the AI does, when, and how humans stay in control)
The right guardrails (provenance, tone, permissions, failure modes)
A practical plan to ship (roadmap, risks, and next-phase build plan)
Ideal for teams who:
have a real workflow to improve (not just “let’s add a chatbot”)
need alignment across stakeholders
want something tangible to test with users quickly
Work
A few selected projects that show how I work – outcome-first, prototype-led, and designed for real constraints.
Family wellness assistant
A real-time vision prototype of a context-aware family assistant – mobile experience + live back-end console for demos, alignment, and testing.
AI-Powered Concierge console
3.5 days → ~2 minutes for proposal creation by restructuring the workflow around a proposal model (“atoms” and “molecules”).
Tallis habits - Web-app for schools
Behaviour design in practice: a learning habits app designed for reflection and motivation over monitoring – co-created with pupils and adopted across the school.
Blog
2025: The year of building
Apple Agent Store – Capability as a Service
Local-first AI
