🔒 Chatbot in e-commerce or a restaurant web page

[tropic mirage]

Hello everyone,
I’m currently learning Python, and I’m interested in building a custom chatbot for an e-commerce web app. I’m aware that there are third-party tools and services that offer chatbot integrations, but I’d prefer to build my own solution without relying on paid subscriptions.

Specifically, I want to create a chatbot that:

Can be embedded on a web page

Uses LLM APIs (like OpenAI, Gemini, etc.) or prompt engineering techniques

Has access to and can interact with my website’s specific data (product information, custom support, can do email to the owner, taking orders, FAQs, etc.)

I’m looking for guidance on how to approach this—whether it’s through API connections or any other effective method.

If anyone has experience with this or can point me in the right direction, I would really appreciate your help!

Thank you!

[plucky escarp]

@tropic mirage

Python help channel opened

Remember to:

• Ask your Python question, not if you can ask or if there's an expert who can help.
• Show a code sample as text (rather than a screenshot) and the error message, if you've got one.
• Explain what you expect to happen and what actually happens.

:warning: Do not pip install anything that isn't related to your question, especially if asked to over DMs.

Closes after a period of inactivity, or when you send !close.

[sharp mist]

product information, custom support, can do email to the owner, taking orders, FAQs

how to approach "product information" will vary depending on how many products you have in total, how often you update that information, how you are storing it, in which way(s) you want to query it etc.
(the simplest way would be a tool which just performs a normal text search over its name, the same way traditional search bars work - that can work well if users are expected to search by the name, as opposed to generic details about the product)

"custom support" you mean customer support? That will vary a lot depending on what you have today and how you want to integrate it
(one way would be just redirecting the user to talk with customer support instead of the chatbot)

"do email" is a bit simpler, just a tool that sends an email may work - just remember to leave the email as a configuration, not as a parameter, not hardcoded

"taking orders" will also vary drastically based on what you already have today, but in a nutshell just create a tool that translates something that makes sense for the model to something that makes sense for your APIs

"FAQs" could be done either similar to product search, or using semantic search Retrieval Augmented Generation

'embedding on a web page' also varies a bit, you could just use an open source frontend inside of an iframe, but ideally you should look for a solution better integrated with whichever framework you're currently using for the frontend (which again, depends on which framework you're using)

I'd recommend

• create a MCP server for the tools, that way you can somewhat freely swap between different agent libraries or connect others proprietary tools
• (in part my personal bias) take a look at pydantic.ai for the agent (chatbot) itself
• use whichever open source chat window component you can find for your frontend framework
• do not over stress prompt engineering until you get the tools done and settle on a LLM model, and adjust the tools as much as you can before you adjust the prompt

for the website specific part, I cannot recommend much else without knowing things specific to your website

[odd charm]

What platform/framework are you using for your website? E.g. Wordpress or Drupal or Joomla or Shopify or Magento or some other custom built e-commerce framework?

[plucky escarp]

Python help channel closed for inactivity

This help channel has been closed. Feel free to create a new post in #1035199133436354600. To maximize your chances of getting a response, check out this guide on asking good questions.