Light the way from idea to launch.
The open-core AI-powered UX research platform that transforms messy startup ideas into validated product blueprints.
Demo · Documentation · Discussions · Roadmap
ShepLight helps non-technical founders go from "I have an idea" to "I know exactly what to build first" in minutes, not months.
We believe every great product starts with understanding your user deeply. ShepLight combines AI-powered research, UX methodology, and actionable frameworks to guide founders through the critical early stages of product development.
- Women in tech who are tired of being overlooked by generic ideation tools and want research-led clarity.
- Non-technical founders who need a coach that translates vision into an actionable product brief.
"I have a great idea, but I don't know where to start."
Most founders waste weeks on the wrong features because they skip user research. Hiring a UX researcher costs $50k+. Reading about UX methodology takes months. ShepLight gives you senior UX researcher thinking in minutes.
Transform a messy idea into a crystal-clear problem statement, target user, and value hypotheses. Powered by AI that thinks like a senior product strategist.
Generate realistic user personas, pain point maps, and emotional journey insights. Built on research synthesis, not generic templates.
Get an actionable MVP scope, prioritized feature list, and 4-week roadmap. Every recommendation traces back to user pain points.
Auto-generates a mind map directly from your Compass, Muse, and Blueprint insights the moment you open it. When journey data is missing, the legacy AI-assisted builder is still available on demand.
Save and version your product decisions. Never lose track of what you committed to build.
Export professional Markdown briefs to share with co-founders, investors, or developers.
- Journey-first mind map — Vault and Export links now pass full project context so the mind map instantly renders your real journey data.
- Insight polish — Positioning copy inside the Insight Report uses the correct readable colors.
- Branded sharing — Custom Open Graph and Twitter meta images ensure ShepLight links look professional on social platforms.
- Node.js 18+
- Supabase account (free tier)
- Groq API key (free tier)
- Firecrawl API key (optional, for research)
# Clone the repository
git clone https://github.com/Radix-Obsidian/Shepherd-Insight.git
cd Shepherd-Insight
# Install dependencies
npm install
# Copy environment template
cp .env.example .env.local
# Add your API keys to .env.local
# See docs/SETUP.md for detailed instructions
# Start development server
npm run devOpen http://localhost:3000 and start your journey.
| Layer | Technology |
|---|---|
| Framework | Next.js 14 (App Router) |
| Language | TypeScript |
| Styling | Tailwind CSS |
| Database | Supabase (PostgreSQL + RLS) |
| Auth | Supabase Auth |
| AI Engine | Groq (Llama 3.3, GPT-OSS) |
| Research | Firecrawl |
| State | Zustand |
| Testing | Cypress, Jest |
| Deployment | Vercel |
ShepLight guides users through a structured 3-step journey:
┌─────────────────────────────────────────────────────────────────┐
│ │
│ 🧭 COMPASS 🎭 MUSE 📐 BLUEPRINT │
│ ────────────────────────────────────────────────────────── │
│ │
│ "What problem "Who are my "What should I │
│ am I solving?" users really?" build first?" │
│ │
│ → Problem Statement → User Personas → MVP Scope │
│ → Target User → Pain Points → Feature Priority │
│ → Value Hypotheses → Insights → 4-Week Roadmap │
│ │
└─────────────────────────────────────────────────────────────────┘
Each step builds on the previous, ensuring every feature recommendation is grounded in user research.
- Compass: Idea clarity engine
- Muse: Persona synthesis
- Blueprint: MVP planning
- Mind Map: Visual strategy builder
- Decision Vault: Version control for decisions
- Exports: Markdown briefs
- Firecrawl deep research integration
- Competitor analysis dashboard
- Interview question generator
- Pain point validation scoring
- Team workspaces
- Commenting and feedback
- Shareable public links
- Notion/Linear integrations
- Custom AI model training
- White-label deployments
- SSO/SAML authentication
- Advanced analytics
ShepLight is an open-core project. The core platform is open source, with premium features planned for sustainability.
We welcome contributions! Here's how to get involved:
- 🐛 Report Bugs — Open an issue
- 💡 Suggest Features — Start a discussion
- 📝 Improve Docs — PRs for documentation always welcome
- 🔧 Submit PRs — See CONTRIBUTING.md
# Fork and clone the repo
git clone https://github.com/YOUR_USERNAME/Shepherd-Insight.git
# Create a branch
git checkout -b feature/amazing-feature
# Make your changes and test
npm run type-check
npm run lint
npm test
# Submit a PRSee CONTRIBUTING.md for detailed guidelines.
| Document | Description |
|---|---|
| Architecture | Technical architecture and design decisions |
| Setup Guide | Detailed installation instructions |
| API Reference | API endpoints and usage |
| Deployment | Production deployment guide |
| Contributing | Contribution guidelines |
- GitHub Discussions — Ask questions, share ideas
- Issues — Report bugs, request features
ShepLight is open source under the MIT License. See LICENSE for details.
Built with love by Golden Sheep AI and powered by:
- Next.js — React framework
- Supabase — Backend infrastructure
- Groq — AI inference
- Firecrawl — Web research
- Vercel — Deployment
- Tailwind CSS — Styling
Made with 🐑 by Golden Sheep AI
"Start with the customer experience. Work backwards to technology." — Steve Jobs