On a WhatsApp message via Evolution API, runs a Gemini agent with Redis and Postgres memory to reply conversationally.
On a form submission, runs Google Gemini over aggregated input, emails results via Gmail, and indexes documents into Pinecone for chat-based retrieval.
On manual run, fetches and extracts web content with Google Gemini, then embeds it and upserts the vectors into a Pinecone store.
On a webhook query, answers from WordPress content embedded into a Supabase vector store, using OpenAI and Postgres chat memory.
RedisGoogle Gemini
Postgres
