Product Setup & Training Your AI
The 4-Step Setup Wizard
After creating your organization, you need to create your first product. This is a guided 4-step wizard that sets up your product, trains the AI to understand it, and identifies your market competitors.
Step 1 — Name & Details
Name Your Product
Provide your product's name, description, industry, and target audience. This helps the AI understand the context of your product from the start. Be specific — describe what your product does and who it's for.
The product name is also used to generate unique feature keys (e.g., AMP-15 for a product called "Amplify"). Choose a name that creates clean, recognizable prefixes.
Step 2 — Feed the AI Context
This is the most important step. You're teaching the AI about your product by providing source materials. Three input types are supported:
🌐 Web URL
Paste your product's website URL, documentation, or landing page. The AI will crawl the page, extract content, and generate a summary. You can optionally enable website crawling to follow links and process multiple pages (up to 30).
📄 Upload PDF
Upload PDF documents like pitch decks, product specs, or design documents. The AI extracts text and generates a summary.
✏️ Manual Text
Paste text content directly. Useful for internal notes, product briefs, or content from other tools that isn't available as a URL or PDF.
Aim for at least 3-5 quality context sources before relying heavily on AI features. The more the AI knows about your product, the better its recommendations, generated PRDs, and story writing will be.
What Makes a Good Context Source?
- Your product website (landing page + features page)
- Product documentation or help center
- Pitch deck or investor materials
- Internal product briefs or strategy docs
- Feature specifications or design documents
Step 3 — Review Product Summary
AI-Generated Summary
The AI processes your context sources and generates a Product Summary (also called the Master Profile). Review and edit this summary before finalizing. Once you confirm, the platform builds your AI intelligence and prepares for discovery.
Step 4 — Competitor Discovery
Discovery Engine
In the final step, the MIND Discovery Engine scans the web, app stores, and industry databases to find your competitors. This typically takes 60-90 seconds. You can then review the suggestions and approve the ones you want to track.
The AI doesn't just look for names; it analyzes product offerings and ranking criteria to find tools that truly compete with your unique value proposition. See the Competitor Discovery Guide for more details.
The Product Summary serves as persistent context for all AI features across the platform — chat, feature generation, PRDs, story writing, and scoring. It's the foundation of MindBacklog's intelligence.
After Setup
Once you confirm your competitor selections, you're routed directly to your Competitive Intelligence page — not a blank dashboard. While the channels for your tracked competitors are being built in the background, you'll see a gathering state that transitions into the full gap matrix and market-signal feed as soon as initial data lands. This is the post-onboarding "wow moment": you set up the product five minutes ago and you're already looking at real competitive intel.
From there:
- Your subscription starts (14-day free trial with full access).
- A guided tour may appear highlighting key areas of the interface — you can dismiss it at any time.
- Ask Mind shows a contextual starter prompt seeded from your product context — one click to start exploring.
- New products default to the RICE prioritization framework (Reach × Impact × Confidence ÷ Effort). You can switch to WSJF at any time in Product Settings.
- To add more products later, use the product switcher in the sidebar.
- Each product has its own independent data, context, features, and settings.
Keeping Your AI Sharp
Your initial context sources are just the starting point. As your product evolves:
- Add new sources when you launch features, publish documentation, or refine positioning.
- Merge staged sources into your Product Summary regularly (The Mind → Knowledge → Merge Staged Sources).
- Review the diff — the AI shows additions, conflicts, and updates before you approve.
- Manually edit the Product Summary if needed — previous versions are tracked for rollback.
A stale Product Summary means stale AI recommendations. If your product strategy shifts, update your context sources and re-merge. The AI automatically filters out irrelevant content, so don't worry about adding too much.