AutoAgent was originally built using no-code tools like Voiceflow and Zapier to quickly test real-world usage and validate product-market fit. After early feedback, we shifted to a more scalable architecture using Lovable.dev (for the frontend widget), n8n (for backend logic and LLM-based recommendation flow), and Supabase (for authentication and chat history). This low-cost, flexible stack helped us move fast without a technical co-founder. Beyond tech, the product was shaped closely with feedback from D2C founders. We sat in on live sales calls, analyzed customer chat transcripts, and mapped decision flows to mimic the experience of a human product expert. Our goal was not just to build a chatbot, but a digital sales assistant that knows when to listen, when to recommend, and when to convert. Our early pilot was with a skincare brand called Tatved, where we saw a 16% revenue bump and a 5% increase in AOV. We're now preparing for a broader launch with 3 brands in our early pipeline. Competitors include Verifast.ai and Shopify’s AI initiatives, but what sets us apart is the level of insight we provide to founders, like top-searched queries, user device data, and soon, customer profiling. Our vision is to democratise AI-powered sales assistants for every Indian D2C brand, starting with plug-and-play deployment and growing into full-stack personalisation infrastructure.
