Author(s)
Andhare Yash Amar, Mane Yashwant Bhaskar , Kharat Pranav Santoshrao, Prof. Komal Palikondawar
- ISSN (P): 3139-8464
- Manuscript ID: 141137
- Volume: 2
- Issue: 10
- Pages: 670–682
Subject Area: Computer Science and IT
Abstract
Most startups fail early, often because ideas were never validated in a structured, unbiased way. Founders rely on gut feeling, friends and family, or costly consultants. This work proposes a platform that combines NLP, ML classifiers, LLM reasoning, and live external data (market trends, startup and funding databases). It takes a text description of an idea and returns a feasibility report. The report has five sub-scores (market demand, competition, originality, technical feasibility, revenue potential), a composite score, a SWOT analysis, and improvement suggestions. The goal is to cut validation time from weeks to minutes for students, first-time founders, and incubators.