Linear Search

• 

  Simple reddit search (App)

• 

  Landing page

• 

  Create validation for an idea screen (App)

• 

  Validation screen (App)

• 

  AI research screen (App)

• 

  AI research card (App)

• 

  An idea generated by AI card (App)

• 

  Create MVP Plan from an idea screen (App)

• 

  MVP Plan of an idea generated by AI (App)

• 

  Ideas screen (App)

Inspiration

I’ve always wanted to build something of my own, but I constantly struggled with coming up with the right idea. Every time I tried brainstorming, I’d hit a wall. Then I realized that Reddit is full of real people talking about real problems every day. What if I could build a tool that listens to those conversations and helps people like me discover and validate startup ideas?

What it does

Linear Search is an AI-powered startup research tool that helps you go from “I want to build something” to “I know exactly what to build.”

It does a few key things:

• AI researcher that researches a topic and give summary and insights.
• Finds startup-worthy ideas by analyzing real Reddit discussions and using AI.
• Validates ideas by auto-posting to Reddit and collecting real feedback
• Analyzes competitors and where can we make impact.
• Generates a complete MVP plan — including what to build, estimated cost, tech stack, and development difficulty e.t.c

It’s like having a research partner who never sleeps and reads Reddit all day.

How we built it

I used a few tools to make this come together without building everything from scratch:

• Boltnew for the frontend.
• Supabase as the backend database.
• n8n to build automation workflows for scraping Reddit, running GPT prompts, and handling feedback collection.
• Netlify for hosting — fast deployments and simple setup
• Entri for registering the domain
• GPT-4o-mini to handle everything from analyzing Reddit threads to writing posts and MVP plans

Challenges we ran into

• Integrating database
• Building n8n backend
• Code got bloated, so AI was hallucinating too much, Had to start from scratch again.

Accomplishments that we're proud of

• It's my first web app which i built and hosted from scratch without knowledge of either frontend or backend.

What we learned

• Some tips on how to prompt AI i.e. in clear step by step manner
• Ask AI to split code in smaller parts. It gives better results and do not hallucinate as much and consumes less tokens

What's next for Linear Search

• Polishing the app more, adding other sources than reddit like quora, web search e.t.c.
• Better competitor analysis from sources like product hunt.
• Improve AI.
• Adding more useful features.

Built With

• boltnew
• github
• n8n
• supabase