How it should work
One conversation
Natural-language prompts move the user from project setup to structured artifacts without restating the same system each time.
Live Demo: Stage 1
Project setup and first generation, shown step by step.
This separate demo page shows the first five minutes of the ShopAI.ca experience in detail: sign-in, project creation, the first prompt, the generated result, and the next prompt-ready workflow for the staff.
The scenario is a Canadian sports game-tape analysis platform with PIPEDA, role-based access, and WCAG 2.1 AA enabled from the start. Sarah gets three usable requirements in minutes while the system keeps every artifact aligned to the same project context.
The demo proves how ShopAI.ca thinks before it proves how much it can generate.
Complete Flow Summary
The full 60-minute path stays visible while Stage 1 gets the detailed treatment.
This page zooms in on the first requirement-generation loop, but the larger flow stays intact from project creation through export.
Time
Action and output
0 min
Create project and store base context for industry, stack, and Canadian compliance.
5 min
Generate FR-001 from "tag defensive breakdowns in transition" and establish the requirements pattern.
10 min
Generate FR-002 for clip search and filtering, linked to FR-001.
15-50 min
Move through use cases, data models, and deployment with context-aware output changes.
60 min
Export a runnable package with code, docs, migrations, infrastructure, and optional GitHub push.
How it will work
Context + routing
The prompt system changes output by view, prior artifacts, compliance, and urgency as the project moves forward.
Where this goes next
See the full 2026 demo
Open the broader 2026 walkthrough for the later stages beyond requirements.
Part 1
Sarah signs in and creates the project.
The first screens are simple on purpose. They gather the metadata that every later prompt will depend on.
What Sarah sees
Authentication, dashboard, and project form.
Sarah signs in with Clerk, lands on an empty dashboard, clicks New Project, and fills in the game-tape analysis form with Node.js, React, PostgreSQL, Docker, AWS, and the three compliance checkboxes.
- Clerk handles authentication before any project data is created
- The dashboard is intentionally empty for first-run clarity
- The form captures enough structure to drive later prompt behavior
UI walkthrough
First-run flow
ShopAI.ca
Canadian AI Systems Design Platform
[Sign in with Clerk]
Dashboard
Welcome, Sarah!
Your Projects: (None yet)
[New Project] [Learn Demo]
Create New Project
Project Name: Game Tape Analysis Platform
Industry: Sports / Video Analysis
Description:
Review game tape, tag possessions,
detect breakdowns, and generate coach notes
Tech Stack:
[x] Node.js [x] React [x] PostgreSQL
[x] Docker [x] AWS
Compliance:
[x] PIPEDA
[x] Role-based access
[x] WCAG 2.1 AA
[Create Project] [Cancel]
Part 2
The admin login is where the private controls and project assets live.
The public demo stays focused on prompts and outcomes. The private admin workspace is where project controls, access rules, and export settings live.
What lives in admin
Project control without exposing the build details.
Once the project is created, the private admin login holds the project list, prompt history, artifact library, access roles, and export controls for the analysis team.
- Projects and uploaded games are organized in one workspace
- Prompt history stays tied to each analysis project
- Access and export controls stay out of the public-facing demo
Admin login
Private controls
Admin Login
Project
Game Tape Analysis Platform
Workspace
- Prompt history
- Artifact library
- Staff roles
- Export controls
- Review status
Access
- Analyst
- Coordinator
- Coach
Latest activity
- Project created
- First prompt ready
- Demo workspace active
Part 3
The editor loads with Requirements selected by default.
Sarah is dropped directly into the first productive view, with the prompt box already tuned for functional requirement generation.
What Sarah sees
Project editor with context-aware prompt box.
The left sidebar shows the system views. The main content area displays a context badge, the requirements prompt, and tips for the next capabilities Sarah can try.
- Requirements are the default entry point
- The placeholder text adapts to the active view
- The UI is simple because the context system carries the complexity
Project editor
Requirements view
ShopAI.ca proj_67890abc
LEFT SIDEBAR MAIN CONTENT
Requirements <- CURRENT Project: Game Tape Analysis Platform
Use Cases Context Badge:
Interactions FUNCTIONAL REQUIREMENTS
Data Models
Architecture Smart Prompt Box:
Deployment Type a capability:
Export e.g. "tag breakdown"
[tag defensive
breakdowns in transition]
[Generate]
Part 4
Sarah enters the first prompt and sees the result take shape.
The point of the demo is the visible result: one focused prompt becomes a structured requirement that the team can keep using.
What Sarah types
The first prompt sets the analysis pattern.
Sarah types "tag defensive breakdowns in transition" and the system turns it into a requirement the staff can review, refine, and build on.
- The prompt is short and written in coaching language
- The result stays specific to game tape and analyst workflow
- The next prompts can build directly from this first artifact
Prompt and result
FR-001
Prompt
"tag defensive breakdowns in transition"
Result
FR-001: Defensive Breakdown Tagging
Description
System identifies and tags each defensive breakdown clip.
Acceptance Criteria
- Tag quarter, clock, possession, and lineup
- Mark missed rotations and late closeouts
- Allow analyst review before save
- Link each flagged clip to a staff note
Part 5
The generated result appears as a card the staff can keep using.
Sarah sees the requirement arrive in a format that is ready for review, follow-up prompts, and later export.
What Sarah sees
FR-001 appears in about 15 seconds.
The first response becomes a complete requirement with acceptance criteria, priority, related compliance notes, and linked next actions. The UI then offers next steps and links to related artifacts.
- Streaming gives visible progress instead of silent waiting
- JSON is validated before it becomes a saved artifact
- Saved artifact history makes the next prompt smarter
Generated result
Requirement card
Generated Requirement
FR-001: Defensive Breakdown Tagging
Description
System tags each defensive breakdown clip from the uploaded game tape.
Acceptance Criteria
- Flag late closeouts
- Flag missed switches
- Tag quarter, clock, and possession
- Attach one analyst note per clip
Actions
- View full requirement
- Generate use case
- Export artifact
Part 6
The next prompts get better because the first artifact is already in context.
Sarah does not restart the system. She keeps typing, and the workspace keeps adding structure.
What happens next
FR-002 and FR-003 build on FR-001.
Clip search and coaching-note workflows are generated in the same pattern, but the system now understands the earlier tagging requirement and starts linking artifacts and compliance notices automatically.
- FR-002 references FR-001 for future tagging and search logic
- FR-003 adds staff-note and cutup workflow dependencies immediately
- The dashboard starts tracking artifact count, status, and progress
Progress after 5 minutes
What Sarah has
Generated Requirement #2
FR-002: Clip Search & Filtering
- Search by quarter, possession, and player
- Filter by tag, coverage shell, and result
- Jump directly to matching clips
- Save analyst views for later review
- Export filtered clips to a cutup
Generated Requirement #3
FR-003: Coaching Notes & Cutups
- Attach staff notes to tagged clips
- Build cutups by theme or opponent
- Share selected reels with role-based permissions
- Preserve timestamps and possession links
- Export notes with clip references
Project Overview
Requirements: 3
FR-001
FR-002
FR-003
Progress: 20%
Next steps:
1. Use Cases
2. Data Models
3. Architecture
Why It Matters
What Sarah gets in 5 minutes used to take days.
This is the simplest way to explain the value of the system: same outcome category, radically different timeline and working cadence.
Before ShopAI.ca
With ShopAI.ca
Day 1-2: schedule and run architect interviews.
Minute 0: create project with stack and compliance rules.
Day 3-5: draft initial requirements and review them with stakeholders.
Minute 1-3: generate FR-001, FR-002, and FR-003 in sequence.
Day 6-10: revise, estimate, and prepare a handoff package.
Minute 4-5: review the generated set and move directly to use cases.
Manual first rounds usually involve interviews, revisions, and handoff delays.
The demo moves directly from project setup into a structured requirement set and next-step flow.
Speed
Three requirements in five minutes
FR-001, FR-002, and FR-003 are ready before a normal game-review requirements session would even be booked.
Quality control
Structured and traceable
The output arrives in a consistent format with clear next actions, linked artifacts, and staff-ready review flow.
Canadian fit
Compliance from the first prompt
PIPEDA, role-based access, and accessibility rules shape the result before code or infrastructure are generated.
Call to action
Use the separate live demo when you want to show the system in motion.
This page is built for walkthroughs with technical buyers who want to see the first generation loop in detail before moving into the broader 2026 platform story.
Book the Live Demo
Walk through the Stage 1 loop with a real project and live prompts.
Discuss Your System
Map your own project views, compliance rules, and artifact flow into the same pattern.
Talk to ShopAI.ca
Use the live demo as the starting point for a deeper build or sector-specific implementation.
Separate Demo
ShopAI.ca
Live walkthroughs for Canadian AI system design.
What this page proves
Powered by ThinkStart.ca
The first requirement loop is enough to show the whole model: project context in, structured artifacts out, with validation, storage, and future retrieval built in.