Describe what your assistant should do. The estimator recommends the type of AI system you need and the build effort — the two numbers that decide what a chatbot really costs.
Fill in the fields — results update live.
Updates live as you type.
An AI assistant has a one-off build cost (design, knowledge preparation, integrations, evaluation) and a running cost (model usage, hosting, WhatsApp conversation fees, and ongoing review). Model usage grows with conversation volume and answer length, so high-volume assistants route simple questions to smaller, cheaper models.
Most of the accuracy of a RAG assistant comes from clean, current, well-structured source content — not the model. Budget time to consolidate FAQs, policies, and product data before launch.
A scripted chatbot follows fixed menus. A RAG assistant retrieves passages from your documents and answers with citations. An AI agent can also take actions — bookings, order lookups, ticket creation — through connections to your systems.
A RAG assistant on a single channel typically takes 3 to 6 weeks including evaluation. Multi-channel assistants and agents that act in other systems take 6 to 12 weeks or more.
Yes. The same knowledge and logic can serve several channels; each channel adds interface, compliance, and testing work.