From discovery to delivery: making sure AI builds the right things

This is the seventh blog in Mike’s 8-part series about how AI can become a force for clarity and acceleration across the broader software development lifecycle.

The delivery loop connects product intent to what engineering actually builds. Stories are synchronised to an issue tracker via /jira, preserving traceability between markdown artifacts and tickets. Story IDs are applied as labels, acceptance criteria populate descriptions, and dependencies are represented as linked issues.

While implemented with Jira in this instance, the approach is tool-agnostic. Story files remain the source of truth; trackers provide alternative views for teams with different preferences, leadership visibility, or automation needs.

The delivery loop

This article is intentionally not covering the engineering parts of the delivery lifecycle to focus on product management activities, so it’s represented here by a literal black box.

Recording releases

After deployment, the /rel command analyses git history to identify delivered work. Commits are matched to stories where possible, and unmatched changes e.g. bug fixes, are flagged for review.

Once confirmed, the command updates story status, removes completed stories from the backlog, and generates release documentation. It updates the product specification, keeping shared context current, ready for subsequent work.

The as-built product spec

The product specification records what is actually running in production: delivered capabilities, architectural characteristics, and known limitations.

When the LLM references this specification, it avoids re-specifying existing functionality and can reason accurately about integration and extension. Where appropriate, screenshots of live systems can be generated and linked.

This artefact supports LLM-assisted discovery and delivery, living alongside code in a concise, structured format.

“By maintaining this shared, up-to-date context, teams ensure that each subsequent cycle of discovery and delivery is informed by an accurate picture of what already exists. Together, these practices reinforce a tight feedback loop between discovery, delivery, and learning.”

Historically, maintaining documentation reflecting the current state of software is a difficult task. AI can not only take away this burden, but leverage the resulting artefacts in a feedback loop to accelerate and improve the quality of downstream enhancements.

Read the full Evolving product discovery for AI delivery eBook

About the author

Mike Mitchell is a Product and Strategy Principal at Equal Experts, helping organisations adopt modern product practices and integrate AI into their product lifecycle in practical, sustainable ways.

His current focus is on helping product teams do more meaningful strategic work with AI, both by accelerating the routine work that crowds it out, and by making AI a more proactive, multi-modal collaborator.

Mike brings 35 years of experience in enterprise software, with a foundation of engineering followed by two decades of product leadership. He is a founder of multiple startups, has advised dozens more, and has led multiple enterprise transformations. Just prior to EE, Mike served as a fractional Chief Product Officer, guiding early- and growth-stage firms to deliver value faster through stronger product strategy and operating models. He holds a degree in Computer Science & Engineering from MIT.

Mike’s final blog in this series pulls together his thoughts about EE’s overall approach to AI-enabled discovery and specification, “…providing the stability needed to adopt more autonomous capabilities without relinquishing control over outcomes that matter.”