Thodar AI
An AI-powered sales assistant tailored for e-commerce, designed to understand customer intent and provide contextual product recommendations.
THE BUSINESS PROBLEM
E-commerce platforms are good at searching products, but customers don't always know what to search for.
They describe a requirement in natural language, while the system expects keywords.
That gap creates friction.
THE IDEA
Thodar AI turns product discovery into a conversation.
A customer can describe what they need, and the system converts that requirement into relevant product recommendations.
THE BUSINESS & PROCESS ANALYSIS
I mapped the buying journey: Need → Search → Compare → Decide.
The biggest friction was at search, where customers had to translate their needs into product-friendly keywords.
THE REQUIREMENTS & IMPLEMENTATION
Key requirements included intent detection, requirement extraction, product matching and conversational recommendations.
The prototype used Next.js, TypeScript, Tailwind CSS and Gemini.
THE RESULT / CURRENT STATUS
Thodar AI reached the prototype stage as an AI-powered sales assistant.
The project showed how AI can reduce the gap between what customers mean and what systems expect.
WHAT I LEARNED
The best AI experiences don't make users adapt to the software.
The software adapts to the user.