Give buyers quick answers while your team focuses on closing deals. A chat assistant greets each user, searches products by city and state, and brings facts from a knowledge source. It fits teams that sell many plans or SKUs and need fast, accurate replies.
A public chat entry starts the flow. An If check and Set nodes prepare an intro message that stores user context into a Postgres table with a short memory window. The main OpenAI assistant runs with a longer Postgres memory and can call three tools. It queries MySQL for matching products using filters like cityQuery, state, modality, and holder count, reads a knowledge base URL for extra info, and posts to an external service with name and birth date to validate records. The assistant then writes a clear answer back to the chat.
You will need an OpenAI API key, plus access to PostgreSQL and MySQL. Create or confirm the aimessages table for chat memory, and check that your Products table and column names match the query. Expect faster replies, less manual lookup, and better qualified leads on your site or in an internal sales console.