AI-Driven Product Discovery and Strategy ( AI-PDS)

AI-driven product life cycle (AIPLC )implementation in AI Desktop Assistants

AI-PDS (AI-Driven Product Discovery and Strategy) is a capabilities-first framework that helps non-technical business leaders — PMs, VPs of Product, CTOs exploring adjacencies — discover, prioritize, and prototype AI use cases before committing engineering resources. Multiple implemenations of the AI-PDS exists for AI driven Product Life Cycle e.g. Agentic IDE, Amazon Quick, Quick Desktop etc,along with prototype creation.

AI Product Discovery and Strategy (AI-PDS) is a structured approach where AI orchestrates every phase from customer discovery to market launch.

Humans retain decision authority at each gate, ensuring alignment with business goals.

This repo provides the Desktop Assistants implementation. For Agentic IDEs implementations please refer to AI-PDS - Agentic IDE implementations of AI-PLC aws-samples/sample-ai-plc

What This Skill Does

Guides Product Managers and business leaders through the AI-PLC Discovery workflow:

Capability	What happens	You get
Envision	Pain point analysis + PR/FAQ (Working Backwards)	PR/FAQ document
Use Case Intake	Document & categorize N use cases	Use Case Repository
Prioritize	Dual scoring (Agentic vs Application)	Ranked Backlog
Prototype	Generate specs + optional HTML prototype	PROTOTYPE-*.md files
Product Strategy	Positioning, business model, KPIs	Strategy Document
Go-to-Market	Segmentation, sales motions, launch	GTM Plan

Three entry points: Start from pain points, use cases, or existing prototype specs.
Every stage is an exit point: Stop whenever you have what you need.

---

Supported Implementations

This repo currently supports:

Implementation	Setup
Amazon Quick Desktop	Default. See instructions below
Claude Desktop	See claude-desktop/README.md for setup instructions.

More desktop assistants coming soon — stay tuned.

---

Get Started with Quick Desktop

1. Open Agents & skills from the left sidebar
2. Click the Skills tab
3. Click + Create → Import URL
4. Paste this repo URL: https://github.com/aws-samples/sample-ai-pds-desktop
5. In any conversation, say: "Start AI-PLC"
6. Follow the guided workflow

For Claude Desktop setup, see claude-desktop/README.md.

---

File Structure

ai-plc-skill/
├── SKILL.md                              ← Orchestrator (main entry point)
├── README.md                             ← This file
├── rules/
│   ├── common/
│   │   ├── process-overview.md           ← Capability model & principles
│   │   ├── session-continuity.md         ← State management & resumption
│   │   └── question-format-guide.md      ← Decision card formatting rules
│   └── discovery/
│       ├── envision.md                   ← Pain points + PR/FAQ
│       ├── solution-analysis.md          ← Single vs. multiple solutions
│       ├── use-case-intake.md            ← Structured use case documentation
│       ├── use-case-prioritization.md    ← Dual scoring frameworks
│       ├── prototype-context-generation.md ← Spec file generation
│       ├── prototype-building.md         ← HTML prototype creation
│       ├── product-strategy.md           ← Strategy development
│       └── go-to-market.md              ← GTM planning
└── templates/
    ├── prfaq-template.md                ← Working Backwards PR/FAQ
    ├── use-case-card.md                 ← Use case documentation format
    ├── scoring-agentic.md              ← Agentic scoring framework
    ├── scoring-application.md           ← Application scoring framework
    └── prototype-spec-template.md       ← PROTOTYPE-*.md template

---

Cross-Platform Compatibility

The PROTOTYPE-*.md files generated by this skill use the same format as the Kiro and Claude Code implementations. A spec generated in Quick Desktop can be handed off to:

• Kiro (for full IDE-based prototype builds)
• Claude Code (for terminal-based prototype builds)
• Use in Desktop tools like Quick Desktop workflow as is for prototype creation

And vice versa — PROTOTYPE-*.md files from Kiro/Claude Code can be used in Quick Desktop (Entry Point 3).

---

Workshop Mode (Mob Discovery)

For facilitated workshops:

1. Facilitator runs Quick Desktop and screen-shares
2. Core team contributes verbally → facilitator types
3. AI synthesizes and prioritizes → group validates
4. PROTOTYPE-*.md specs distributed to sub-teams via Slack/OneDrive
5. Sub-teams build in parallel (Quick, Kiro, or Claude Code)

---

Source

• Framework: AI-PDS (AI Product Discovery & Strategy)
• AI-PDS - Agentic IDE implementations of AI-PLC aws-samples/sample-ai-plc
• Related Blog: Henry Schein One goes AI-native with AI-PDS

Security

See CONTRIBUTING for more information.

License

This library is licensed under the MIT-0 License. See the LICENSE file.

Version 1.0 | 2026-07-21