Empower Your Journey
Archive Intelligence
1. Start with the role (this matters most)
WeRock.ai should not be a tool. It should be an engine.
Define it like this:
WeRock.ai = the intelligence layer that turns content into action
That means:
- Not just answering questions
- Not just generating posts
- But organizing truth, media, and people into coordinated output
2. Core functions (keep it tight)
You don’t need 20 features. You need 4 that actually work:
1. Archive Intelligence
Turn everything you’ve created into a structured, searchable system.
- Every post, session, event = indexed
- AI can answer:
- “Show me all Iron City Trykes activity”
- “Timeline of April 23 event”
- “All content referencing Girard”
👉 This becomes your source of truth
2. Content Engine
AI generates publish-ready material in your voice
- Posts
- Summaries
- Headlines
- Video descriptions
- Social clips
But key difference:
It pulls from your archive, not generic internet tone
3. Action Layer
This is where most projects fail—you shouldn’t.
WeRock.ai should:
- Turn stories into calls to action
- Turn events into replicable models
Example:
- “Run the 4.25/4 model in another city”
- “Launch a local Iron City Trykes-style event”
👉 This is how it becomes a movement tool
4. Network Coordination
Lightweight, but powerful.
- Track people (nodes)
- Track projects
- Track participation
Not a social network—more like:
A map of who is doing what, where, and why
3. Technical stack (simple + deployable)
Don’t overbuild. You already started correctly.
Frontend
- Next.js (you’re already there)
- Vercel deployment
Backend
- Supabase (database + auth + storage)
AI layer
- OpenAI API (for:
- chat
- summarization
- embedding your archive)
Critical piece most people miss:
👉 Vector database (for memory)
- Supabase pgvector OR Pinecone
This allows:
- “Ask anything about WeRock history”
- Context-aware generation
4. System architecture (keep this mental model)
Think in 3 layers:
Layer 1 — Content (WeRock.tv)
- Posts
- Videos
- Events
- Documents
Layer 2 — Structure (Supabase)
- Tables:
- documents
- events
- people
- sessions
- projects
Layer 3 — Intelligence (WeRock.ai)
- Search
- Generate
- Connect
- Recommend actions
5. First build (do this in order)
Don’t skip steps.
Step 1 — Archive ingestion
- Load:
- your posts
- your conversations
- your event records
- Store in Supabase
- Generate embeddings
Step 2 — Ask WeRock
Simple interface:
“Ask anything about the network”
Examples:
- “What happened April 23?”
- “Summarize Iron City Trykes”
- “Who is involved in WeRock?”
Step 3 — Generate post
Button:
“Turn this into a publish-ready post”
Step 4 — Timeline view
- Visual archive
- Click → expand → AI summary
Step 5 — Action prompts
- “What should we do next?”
- AI suggests:
- events
- posts
- outreach
6. Guardrails (important for your situation)
You’re dealing with real people + serious claims, so:
Always:
- Label content:
- Primary
- Testimonial
- Unverified
- Use neutral phrasing:
- “X states…”
- “Records indicate…”
👉 This protects credibility and keeps the system usable long-term.
7. Positioning (this is your edge)
Most AI platforms are:
- entertainment
- productivity
- automation
WeRock.ai should be positioned as:
Truth + Culture + Action Engine
Or even tighter:
“From signal to action.”
8. What this becomes (if you execute right)
Not a website.
Not a chatbot.
But:
- A living archive
- A content generator
- A coordination system
- A movement infrastructurent tool