Functional LLM Chatbots - HTMX, Function Calling & LLama 3
06-06, 11:20–12:50 (Europe/Madrid), Workshop

Companies crave LLM integrations. Have you built one in Django?

This workshop provides a step-by-step guide on how to let LLMs take control of your Django application, both on the server-side and the client-side.


{% extends "abstract.html" %}

{% block repository %}
Link to the GitHub repository: https://github.com/scriptogre/functional-chatbots

Includes useful patterns you can reuse.
{% endblock %}

{% block content %}

This workshop is a practical guide to creating dynamic Django applications that are deeply integrated with LLM models using HTMX.

Participants will learn how LLMs can trigger UI changes that are immediately visible to the user, handle CRUD functionality, or any other server-side tasks - in an elegant and concise manner.

The session offers hands-on experience and valuable tips for efficient LLM integrations. You will leave with the knowledge and resources necessary to implement a LLM in your own Django project.

{% block steps %}
1. Sneak peak at final result & cloning the GitHub repository (2 min)
2. Get you up to speed (10 min)

{% if attendees.care %}

  1. Explain 2 methods of rendering chat messages dynamically with htmx (2 min)
  2. Implement (one of) them yourself (10 min)

{% endif %}

  1. Explain how the LLM can modify the UI state in a session (2 min)
  2. Implement it yourself (10 min)
  3. Explain how the LLM can trigger CRUD functionality (2 min)
  4. Implement it yourself (10 min)

{% endblock steps %}

{% block questions %}

I will be free to discuss anything about function-calling, possible security practices against prompt injection, and tools for prompt engineering at the end of the workshop. The very limited time doesn't permit it during the workshop.

{% endblock %}

{% endblock content %}

Video: https://youtu.be/jtjMR47AAd8


Topics

Async, htmx, Rest APIs, HTML/CSS/JavaScript

Audience Level

Intermediate

Hi, I'm Chris.

I've been learning Python for six years and Django for four. I currently work for a team in Albuquerque that focuses on drug discovery and drug repurposing using LLM technology.

My work involves:
- LLM integration (e.g., prompt engineering)
- Building a Django web app to host these technologies (full stack development)

Combining these areas is a niche field where I've gained some interesting insights that I believe are worth sharing.

I've spent hundreds of hours preparing this workshop for you. If you learn something new and useful, the effort will have been worthwhile.

If you'd like to learn more about me, please come and say hi. I enjoy meeting new people.