Hey everyone,
I’m currently working on creating simple customer support agents and lead generation bots, with a focus on European-based businesses.
I’ve been using Voiceflow + Voiceglow (Tixagents) + Make, but I’m exploring my options for integrating with WhatsApp and need some advice.
Here’s what I need to do:
- Build customer support bots that can answer knowledge base questions. [I know how to do]
- Implement lead gen via Airtable integration. [I know how to do]
- The client wants this to run on WhatsApp. [I have never done it before and need advice]
- The bot should be able to respond to general public questions, but also provide “private access” for customers.
- The bot should be simple initially but have the potential to scale and be duplicated for multiple clients.
Here are my questions:
I know Voiceflow with Voiceglow can connect to WhatsApp module
I know n8n has an agent module, and i can connect Voiceflow with voiceglow to it and then route it through WhatsApp
A. Make has its own agent module. Can I connect this to WhatsApp and use it alongside Voiceflow?
B. In terms of performance and cost, if you had to develop and duplicate a simple agent for multiple clients, which option would you choose: Voiceflow with Voiceglow then route to whatsapp app, Voiceflow connected to n8n then whatsapp, or Voiceflow connected Make Agent then Whatsapp?
C. In terms of cost, is n8n cheaper when hosted compared to Make Agent? (I’m talking about clients who won’t have 10,000 requests per month—most will be small to medium businesses with maybe 100-200 requests max).
D. How would you secure access to the private area with the least risk of hacking? Do you verify the phone number of the person chatting? Or do you ask for their email, check it against your database, and then ask for a password (for example, all students in my training program know this password and can access it)? Anyone with real-world experience here?
E. Finally, I want to know: in your opinion, are the LLM knowledge base questions on n8n (using Airtable or Notion) or Make as effective as Voiceflow for answering knowledge base queries?
In terms of performance, cost, and security, have you tried any of these solutions(1,2,3,4,5)? Any feedback?
Thanks in advance for your insights!