AI-Native Product Discovery with Codex Workshop (in-person)
AI makes it dramatically faster to build prototypes. But building faster does not mean discovering the right product faster.
In this live, hands-on workshop, you'll learn an AI-native product discovery workflow for moving from a real customer problem to a focused experiment. Using Codex, you'll investigate workflows, challenge assumptions, explore competing hypotheses, build the smallest useful experiment, verify what works, and decide what deserves further investment.
You will learn how to use AI agents to make better product decisions; this is not a “build an app with AI” workshop.
Problem → Evidence → Hypotheses → Critical Assumption → Experiment → Verification → Learning → Decision
What You'll Learn
Participants will work through one practical discovery cycle using a real customer or internal workflow:
Frame the workflow. Identify the user, desired outcome, friction, workarounds, handoffs, and available evidence.
Investigate the evidence. Use Codex to separate observations from assumptions, identify evidence gaps, and uncover unanswered questions.
Develop and challenge hypotheses. Explore competing explanations and solutions instead of immediately converging on the first plausible idea.
Find the critical assumption. Determine which uncertainty matters most and what must be true for the opportunity to succeed.
Design the smallest useful experiment. Define a prototype, simulation, analysis, or test that reduces the most important uncertainty.
Build with Codex. Use Codex for investigation, planning, implementation, critique, and testing while keeping product judgment human-owned.
Verify before investing. Define acceptance criteria, examine failures, and determine what the experiment does—and does not—prove.
Assess product defensibility. Look beyond model capability to context, workflow integration, data and feedback loops, tools, evaluations, trust, distribution, and switching costs.
Make the product decision. Use the evidence to choose: Kill, Investigate, Iterate, Build, or Scale.
Who Should Attend
Designed for founders, product managers, AI/product engineers, designers, technical product leaders, and innovation teams.
Coding experience is helpful but not required. Access to Codex is recommended.
What to Bring
Bring a laptop and one real customer or internal workflow you want to improve.
Instructor
Andrew Tsintsiruk tsintsiruk is Founder & CEO of Rohic Inc., Managing Partner at Data Strategy Lab, and an AI workshop instructor with The Upskilling Labs. He has more than 15 years of experience across technology, product development, data strategy, customer discovery, and AI-enabled product delivery.
Andrew builds AI-native products and agentic workflows with a focus on multi-agent orchestration, tool calling, retrieval-augmented generation (RAG), evaluation, and human-in-the-loop controls.
Format & Materials
Format: Live, hands-on workshop
Duration: 1.5 hours
Materials included: AI-Native Discovery Notebook, hypothesis templates, experiment canvas, verification checklist, defensibility framework, Codex workflow prompts, and decision rubric.
Bring a real workflow. Leave with evidence, an experiment, and a sharper product decision.