Design & Delivery

The Design and Delivery phases are where your product takes shape. You move from high-level architecture decisions to implementation-ready user stories — all while maintaining a clear link back to the goals and research you established in Discovery and Definition.

This guide walks you through every step: creating product architecture, building story maps, writing user stories, and defining your minimum viable scope. Along the way, Audra Flow's AI agents help you make better decisions and move faster.

Design Phase

The Design phase is the bridge between “what are we building?” and “how will it work?” Here you create the structural blueprint for your product, defining the major building blocks and how they relate to each other.

During Design, you will typically produce two kinds of artifacts:

• Product Architecture — the high-level breakdown of your product into application areas, components, and their relationships.
• Application Area Definitions — detailed descriptions of each major section of your product, including what it does, who uses it, and what capabilities it provides.

Everything you create in the Design phase is linked to the personas, journeys, and specifications from your Definition phase, so you always know why each architectural decision exists.

Building Product Architecture

Product architecture is how you organise your product into logical areas. Think of it as a map of what your product does, broken into sections that make sense for your users and your team.

How to Create Architecture Artifacts

1. Open your project and navigate to the Design phase in the sidebar.
2. Click Add Architecture Artifact. You can create one from scratch or let the AI Architect agent generate a starting point based on your specifications.
3. Give each application area a clear name and description. For example, if you are building an e-commerce platform, you might have areas like “Product Catalog,” “Shopping Cart,” and “Order Management.”
4. Link each application area to the relevant specifications from your Definition phase. This connection is what powers Audra Flow's traceability — you can always trace a piece of architecture back to a specific product need.
5. Review and refine. The Product Guru agent can analyse your architecture for gaps, conflicts, or missing areas.

Tips for Good Architecture

• Keep application areas at a level your whole team can understand. Avoid going too granular too early.
• Make sure every specification from your Definition phase is covered by at least one application area. The Product Guru will flag any that are missing.
• Use the description field generously. Explain not just what an area does, but who it serves and why it matters.

Delivery Phase

Once your architecture is in place, the Delivery phase turns those structural decisions into actionable work. This is where you create the artifacts your team needs to start building:

• Technical Architecture — detailed technical decisions that support each application area.
• Story Maps — visual layouts of user activities and the stories that support them.
• MVP Scope — the minimum set of stories needed for your first release.
• User Stories — the individual pieces of work your team will implement, complete with acceptance criteria.

Each of these artifacts connects back to your architecture, which connects back to your specifications, which connect back to your goals. This chain is what makes Audra Flow's traceability work end to end.

Story Mapping

Story mapping is a way to organise user stories by the activities and tasks your users perform. Instead of a flat list, you see your stories arranged in a two-dimensional map that reveals the big picture.

How Story Maps Are Organised

A story map in Audra Flow has three levels:

Creating a Story Map

1. Navigate to the Delivery phase and select Story Map.
2. Start by defining your activities along the top row. These should align with the application areas from your Design phase.
3. Under each activity, add the tasks that users perform. Think in terms of the steps a user would take to accomplish something.
4. Under each task, add the user stories that describe the specific work needed. You can write these by hand or use the AI Product Owner agent to generate them from your specifications.
5. Drag and drop to reorder stories by priority. Stories near the top are higher priority; stories near the bottom are lower priority.

Prioritising with the Story Map

The story map makes prioritisation visual. Draw a horizontal line across the map to separate what goes into your first release from what can wait. Everything above the line is your MVP scope; everything below is a future release.

Writing User Stories

User stories are the fundamental unit of work in Audra Flow. Each story describes a single piece of functionality from the user's perspective, along with the criteria that define when it is complete.

Story Structure

Every user story in Audra Flow follows a consistent structure:

• Title — a short, descriptive name for the story.
• User Story Statement — written in the format: “As a [type of user], I want [an action] so that [a benefit].”
• Acceptance Criteria — a checklist of conditions that must be true for the story to be considered done.
• Linked Specification — the specification from your Definition phase that this story helps fulfil.
• Priority — where this story sits in the priority order.

Creating Stories Manually

1. From the story map or the stories list, click Add Story.
2. Write a clear title and user story statement. Be specific about who the user is and what they are trying to accomplish.
3. Add acceptance criteria. Each criterion should be independently verifiable — someone should be able to look at the finished work and say “yes, this is done” or “no, it is not.”
4. Link the story to the relevant specification. This maintains your traceability chain.

AI-Assisted Story Generation

Audra Flow can generate user stories directly from your specifications. Here is how:

1. Select one or more specifications in the Delivery phase.
2. Click Generate Stories. The AI Product Owner agent will analyse each specification and produce a set of user stories with acceptance criteria.
3. Review the generated stories. You can edit, split, merge, or discard them as needed. AI-generated stories are marked with a badge so you always know their origin.
4. Accept the stories you want to keep. They are automatically linked to the specification they were generated from.

AI-generated stories are a starting point, not a finished product. Always review them with your team and adjust the acceptance criteria to match your specific needs.

MVP Scope

The MVP (Minimum Viable Product) scope defines the smallest set of stories that delivers enough value to be worth releasing. Getting this right is critical — too much scope delays your release; too little scope fails to solve the user's problem.

Defining Your MVP

1. In the Delivery phase, open the MVP Scope view.
2. Review your story map with your team. Identify the stories that are essential for your product to work at all. These are your MVP candidates.
3. Mark stories as In MVP or Post-MVP. Audra Flow tracks these designations across your story map so you can see the boundary clearly.
4. Check your coverage. The MVP scope view shows you which specifications and goals are covered by your MVP stories and which are not. If a critical goal has no stories in the MVP, you may need to reconsider your scope.

Success Criteria

A good MVP scope has these properties:

• Covers the primary use case — a real user can accomplish the core task your product is designed for.
• Is internally consistent — no story depends on another story that is not in the MVP.
• Is testable — you can verify that the MVP works as intended before releasing it.
• Is releasable — the product is complete enough that users will find it useful, even without the features you deferred.

Using AI in Design & Delivery

Two AI agents are particularly active during the Design and Delivery phases. Here is how they help and when to use them.

The Architect Agent

The Architect agent specialises in structural and technical decisions. It can:

• Generate product architecture from your specifications, suggesting application areas and their relationships.
• Review your architecture for completeness, flagging specifications that are not covered by any application area.
• Produce technical architecture artifacts, including suggested technology choices and integration patterns.
• Answer questions about your architecture in the context of your project — for example, “Which application area handles user authentication?”

The Product Owner Agent

The Product Owner agent focuses on stories and scope. It can:

• Generate user stories from specifications, complete with acceptance criteria.
• Suggest story breakdowns when a story is too large to implement in a single iteration.
• Help refine acceptance criteria to be more specific and testable.
• Identify dependencies between stories and suggest an ordering that minimises risk.

The Product Guru

The Product Guru watches across your entire project and provides ongoing analysis. During Design and Delivery, it can:

• Detect gaps — goals or specifications that have no corresponding architecture or stories.
• Find conflicts — stories or architectural decisions that contradict each other.
• Suggest enhancements — improvements to your stories or architecture based on patterns it has identified.

Traceability

Traceability is one of Audra Flow's most powerful capabilities. Every artifact you create is connected to the artifacts that came before it, forming a chain from high-level goals all the way down to individual user stories.

The Traceability Chain

Here is how the chain works across the five delivery phases:

This means you can pick any user story and trace it back to the goal it helps achieve. You can also start from a goal and see every story that contributes to it. This two-way traceability is essential for audits, stakeholder reviews, and making confident scope decisions.

Viewing Traceability

Audra Flow provides a traceability graph that visualises these connections. You can access it from any artifact by clicking the traceability icon, or from the project overview where you can see the full picture for your entire project.

Next Steps

• Once your stories are written and your MVP is defined, head to Releases & Evidence to learn how to package your work into auditable releases.
• Review the AI Service page to understand how the AI agents work behind the scenes.
• Explore Security & RBAC to learn about permissions and who can edit artifacts in each phase.