YC explorer

• 

🚀 Inspiration The idea came from a simple but painful founder problem: "How do I know if my startup idea has already been tried?"

Today, startup founders, accelerators, and investors struggle to find structured data about what sectors are oversaturated, which business models have been heavily explored, and where true white-space opportunities exist. We realized that with AI and accelerator datasets like Y Combinator’s public data, we could finally build a tool to analyze startup patterns at scale — and turn historical data into actionable startup insights.

🔎 What it does YC Explorer is an AI-powered platform that:

Scrapes and structures accelerator data (starting with Y Combinator)

Uses AI models to classify industries, business models, and sectors

Detects market trends, category saturation, and underexplored opportunities

Suggests potential startup ideas based on real data gaps

Provides an interactive dashboard for founders, investors, and accelerators to explore the data and validate ideas more scientifically.

🔧 How we built it Bolt New: We used Bolt New to rapidly scaffold the frontend, backend API, database schema, and workflow logic.

Supabase: Database and edge functions for scraping orchestration.

AI Models: Used OpenAI / Claude for classification, keyword extraction, and clustering of startup descriptions.

Web scraping: Built custom scraping pipelines for Y Combinator's public portfolio data.

Data enrichment: Applied LLM-powered enrichment to extract industries, verticals, and business models.

Visualization: Simple interactive dashboards to explore trends by category and time.

🧱 Challenges we ran into Handling scraping limitations and structuring unstructured text data into clean, AI-friendly formats.

Designing robust enrichment pipelines to accurately classify startup descriptions, which often have vague or ambiguous wording.

Integrating multiple AI prompts to create consistent and meaningful industry classifications.

Ensuring smooth orchestration between scraping, enrichment, storage, and frontend visualization — especially within Bolt New’s environment.

🏆 Accomplishments that we're proud of Extracted and processed data for over 5,000 YC startups in just a few days.

Built a fully functional AI enrichment pipeline that accurately categorizes startups into meaningful industry clusters.

Created a working prototype that founders, accelerators, and investors can use to visually explore trends and idea gaps.

Successfully demonstrated how AI can turn historical startup data into actionable insights.

🎯 What we learned How powerful AI models are when combined with well-structured datasets — they allow for deep pattern recognition that humans would struggle to detect manually.

The importance of precise prompt engineering and multi-step enrichment workflows to maximize AI classification accuracy.

That founders and accelerators are hungry for real, data-driven tools to help with idea validation and scouting.

Bolt New allowed us to iterate incredibly fast on a full-stack AI project.

🚀 What's next for YC Explorer Expand beyond YC to include other accelerators (Techstars, 500 Global, Antler, etc.)

Add deeper business model clustering (e.g. B2B SaaS, marketplaces, API-first models)

Implement time-series analysis to visualize how sectors evolve batch over batch

Build out a full SaaS subscription model for founders, accelerators, and VCs

Explore partnership or strategic interest from accelerators and investors, including Y Combinator

Built With

• api
• boltnew
• javascript
• openai
• tailwindcss