LOCON: A Full-Stack Geospatial Market Analysis and Competitor Intelligence Platform for Physical Business Viability Assessment

Entrepreneurs and small business owners worldwide face a persistent information deficit when evaluating
the financial and market viability of physical retail locations. Traditional market research remains prohibitively
expensive, largely qualitative, and depreciates rapidly as economic conditions shift. This paper presents Locon, a
full-stack, AI-driven geospatial market analysis platform designed to systematically address these challenges through
engineering. Locon integrates real-time web scraping of global cost-of-living indices from Numbeo, spatial
competitor detection via the Google Places Nearby Search API, and deterministic financial modelling rooted in
Purchasing Power Parity (PPI) heuristics to produce a comprehensive viability assessment for any selected
geographic coordinate. The Haversine formula is applied on the backend to enforce strict metric-radius competitor
filtering, replacing the imprecise bounding-box approximations common in consumer mapping applications. A
deterministic 0–100 Success Score is computed by combining competitor density penalties, capital sufficiency bonuses,
and existing competitor reputation scores. Finally, a Generative AI assistant powered by Google's Gemini 2.5 Flash
model contextualises the quantitative output into actionable advisory narratives. Empirical testing across multiple
city configurations—including Mumbai, London, and New York—demonstrates capital projections conforming
within ±12% of verified commercial leasing benchmarks. The platform is deployed in a fully decoupled
architecture with a React 19 + Vite frontend hosted on Netlify and a Python FastAPI backend on Render, backed
by a SQLAlchemy ORM layer supporting both PostgreSQL and SQLite.