Flowise AI Chatbots Creation & Integration
Budget: €30 – €250 EUR
Flowise Ai ChatBot
• Create 100 separate and autonomous chatbots, with different names.
• Each chatbot must answer questions posed by the user on the text of the documents inserted in the knowledge base of the specific chatbot.
• Each chatbot must only access documents in its dedicated database.
• Users must log in to use the chatbot.
• Each user must have a counter of messages/tokens consumed.
• At the end of the month it must be possible to know how much each user has consumed (per charge).
• The system must be integrated into WordPress (iframe or other).
• Everything must be stable, secure and low maintenance.
________________________________________
Components to build:
1. Chatbot System (Flowise AI)
• Use Flowise AI on VPS server.
• Install and configure Flowise AI.
• Create 100 separate Flows
• Each Flow must have:
o Its own separate knowledge base (PDF, Word, Excel file).
o Restricted access via protected parameters or URLs.
________________________________________
2. Document knowledge base
• Each condominium must upload its own files (1,000+ documents).
• Use a connected vector database:
o Recommended: Qdrant, Pinecone, Chroma.
• Each knowledge must be separated from the others (no document mixing).
________________________________________
3. User authentication system (WordPress)
• Users register and log in to WordPress.
• Each user must have an "Alpha Numeric Code" assigned (custom user field).
• When they log in, they must only see the chatbot of their "Alpha Numeric Code". Example: code: "Alfa01" will only see the documents inserted in the Knowledge Alfa01. It will not see the documents, and will not be able to receive answers, on the remaining 99 Knowledgeladge.
________________________________________
4. WordPress integration
• Each WordPress page loads the correct chatbot.
• Integrate via:
o Secured Iframe (easiest option).
o Or with API + WebApp (more complex option, only if necessary).
________________________________________
5. Message/Token Counter
• Each question sent by the user must:
o Record: WordPress User ID + Question sent + Tokens consumed.
• Save this information in a SQL database (PlanetScale, Railway, Supabase).
• Create a basic system to:
o View how many messages each user has sent.
o Filter by condominium / month / total tokens.
________________________________________
6. Reporting/Billing System
• CSV export or internal dashboard for:
o Username
o Condominium
o Number of messages sent
o Tokens consumed
o Estimated monthly cost
Security required
Protection of knowledge bases: no user should see knowledge of another condominium.
Chatbot access only by logged in users.
OpenAI API key protection (not visible frontend).
Regular backup of SQL and vector database.
• Create 100 separate and autonomous chatbots, with different names.
• Each chatbot must answer questions posed by the user on the text of the documents inserted in the knowledge base of the specific chatbot.
• Each chatbot must only access documents in its dedicated database.
• Users must log in to use the chatbot.
• Each user must have a counter of messages/tokens consumed.
• At the end of the month it must be possible to know how much each user has consumed (per charge).
• The system must be integrated into WordPress (iframe or other).
• Everything must be stable, secure and low maintenance.
________________________________________
Components to build:
1. Chatbot System (Flowise AI)
• Use Flowise AI on VPS server.
• Install and configure Flowise AI.
• Create 100 separate Flows
• Each Flow must have:
o Its own separate knowledge base (PDF, Word, Excel file).
o Restricted access via protected parameters or URLs.
________________________________________
2. Document knowledge base
• Each condominium must upload its own files (1,000+ documents).
• Use a connected vector database:
o Recommended: Qdrant, Pinecone, Chroma.
• Each knowledge must be separated from the others (no document mixing).
________________________________________
3. User authentication system (WordPress)
• Users register and log in to WordPress.
• Each user must have an "Alpha Numeric Code" assigned (custom user field).
• When they log in, they must only see the chatbot of their "Alpha Numeric Code". Example: code: "Alfa01" will only see the documents inserted in the Knowledge Alfa01. It will not see the documents, and will not be able to receive answers, on the remaining 99 Knowledgeladge.
________________________________________
4. WordPress integration
• Each WordPress page loads the correct chatbot.
• Integrate via:
o Secured Iframe (easiest option).
o Or with API + WebApp (more complex option, only if necessary).
________________________________________
5. Message/Token Counter
• Each question sent by the user must:
o Record: WordPress User ID + Question sent + Tokens consumed.
• Save this information in a SQL database (PlanetScale, Railway, Supabase).
• Create a basic system to:
o View how many messages each user has sent.
o Filter by condominium / month / total tokens.
________________________________________
6. Reporting/Billing System
• CSV export or internal dashboard for:
o Username
o Condominium
o Number of messages sent
o Tokens consumed
o Estimated monthly cost
Security required
Protection of knowledge bases: no user should see knowledge of another condominium.
Chatbot access only by logged in users.
OpenAI API key protection (not visible frontend).
Regular backup of SQL and vector database.