AI knowledge base
Business Ideas Knowledge Base
A business idea database with search, filters, scores, and source-backed evidence.
Overview
Business Ideas Knowledge Base is my lab for working with ideas. Instead of keeping a loose notes list, every idea has structure, scores, and sources. The interface helps search for directions by solo-founder fit, AI leverage, capital, and B2B fit.
Problem
Business ideas are easy to collect but hard to compare. Without sources and criteria, the list becomes inspiration noise rather than a useful way to decide what to validate next.
Product bet
If an idea has sources, scoring, and fast search, it becomes easier to separate interesting inspiration from a direction worth validating with a product.
What I built
- Public lab with search, filters, sorting, and random idea discovery.
- Idea cards scored by difficulty, solo fit, and AI leverage.
- Detail modal with summary, customer, time to revenue, and source evidence.
- API routes for search, random ideas, and recommendations.
- Data model prepared for sources, timestamps, and MCP access.
AI layer
AI supports organization and recommendation: an idea has evidence, scoring, and profile context instead of being a plain note.
Architecture
Next.js, TypeScript, local data layer, API routes, Tailwind, Radix UI, and a component frontend embedded into this portfolio.
Key decisions
- A usable lab on the site first, without a separate app or auth.
- Sources and evidence as part of the data model, not a side note.
- UI closer to a calm research tool than a dashboard.
- Data and API separated from the portfolio layer so the lab can grow later.
Outcome / evidence
A working public lab on bogumilluc.pl: users can search, filter, draw a random idea, and open source-backed idea details.
What I would do next
Add more sources, improve profile-based recommendations, and build a workflow for saving ideas into validation.