Don’t just shift your bottlenecks upstream.
AI tools have changed how engineers work. Code gets written faster, reviewed faster, shipped faster. But developers are only one part of the team that delivers software to customers, and when product managers, designers, and delivery leads are left out of the transformation, the bottleneck doesn't disappear. It just moves upstream, back to discovery, requirements, and handoff.
This is the problem Josh Guice, Principal Architect at Liatrio, and Dan Grace, AVP of Client Solutions, set out to address in this session: what does an AI-enabled product workflow look like when it spans the whole path from discovery to developer-ready, not just the code?
Their answer is a live, end-to-end walkthrough: from a raw stakeholder transcript, through story writing and acceptance criteria, to BDD test scenario generation. The workflow is LLM-agnostic, toolchain-agnostic, and built as a pattern any delivery team can adopt, not a one-off demo tied to a specific stack.
From raw transcript to structured requirements
The workflow starts before a developer is ever in the room. A product owner takes a raw meeting transcript and uses AI to extract clean, actionable requirements directly from the conversation, turning an hour of stakeholder discussion into a structured starting point for the story.
Context your stories shouldn't have to go without
Requirements rarely live in isolation. The session shows how to pull relevant context from tools like Confluence, SharePoint, and Jira so that by the time a story reaches the engineering team, the background is already baked in, instead of buried in a thread someone has to go dig up.
Prototyping at the speed of conversation
From there, the demo shows a feature taking shape in real time, with AI supporting rapid prototyping and team feedback cycles. The result is a move from initial concept to a developer-ready spec faster than traditional handoffs allow.
Acceptance criteria that generate their own tests
Perhaps the most concrete payoff: Gherkin-style acceptance criteria and BDD test scenarios generated directly from the story, tracing back to Jira and enabling automated test validation of the original acceptance criteria. Requirements and tests stay connected instead of drifting apart after handoff.
Beyond net-new features
The live demo focuses on adding a new feature to an existing application, but the same workflow applies to modernization, migration, and legacy de-risking work. Those scenarios aren't demoed explicitly in the session, but the underlying pattern, structured requirements in, traceable tests out, holds regardless of the delivery context.
Where to start
You don't need to overhaul your whole product process to try this. Simply pick one upcoming story, run the transcript-to-requirements step, and see what a developer-ready handoff looks like when the background work is already done. That's the workflow this session walks through, end to end.
Related Resources
